Justice O’Connor

delivered the opinion of the Court.

This case requires us to decide whether an initial failure of law enforcement officers to administer the warnings required by Miranda v. Arizona, 384 U. S. 436 (1966), without more, “taints” subsequent admissions made after a suspect has been fully advised of and has waived his Miranda rights. Respondent, Michael James Elstad, was convicted of burglary by an Oregon trial court. The Oregon Court of Appeals reversed, holding that respondent’s signed confession, although voluntary, was rendered inadmissible by a prior remark made in response to questioning without benefit of Miranda warnings. We granted certiorari, 465 U. S. 1078 (1984), and we now reverse.

I

In December 1981, the home of Mr. and Mrs. Gilbert Gross, in the town of Salem, Polk County, Ore., was burglarized. Missing were art objects and furnishings valued at $150,000. A witness to the burglary contacted the Polk County Sheriff’s Office, implicating respondent Michael El-stad, an 18-year-old neighbor and friend of the Grosses’ teenage son. Thereupon, Officers Burke and McAllister went to the home of respondent Elstad, with a warrant for his arrest. Elstad’s mother answered the door. She led the officers to her son’s room where he lay on his bed, clad in shorts and listening to his stereo. The officers asked him to get dressed and to accompany them into the living room. Officer McAllister asked respondent’s mother to step into the kitchen, where he explained that they had a warrant for her *301son’s arrest for the burglary of a neighbor’s residence. Officer Burke remained with Elstad in the living room. He later testified:

“I sat down with Mr. Elstad and I asked him if he was aware of why Detective McAllister and myself were there to talk with him. He stated no, he had no idea why we were there. I then asked him if he knew a person by the name of Gross, and he said yes, he did, and also added that he heard that there was a robbery at the Gross house. And at that point I told Mr. Elstad that I felt he was involved in that, and he looked at me and stated, ‘Yes, I was there.’” App. 19-20.

The officers then escorted Elstad to the back of the patrol car. As they were about to leave for the Polk County Sheriff’s office, Elstad’s father arrived home and came to the rear of the patrol car. The officers advised him that his son was a suspect in the burglary. Officer Burke testified that Mr. Elstad became quite agitated, opened the rear door of the car and admonished his son: “I told you that you were going to get into trouble. You wouldn’t listen to me. You never learn.” Id., at 21.

Elstad was transported to the Sheriff’s headquarters and approximately one hour later, Officers Burke and McAllister joined him in McAllister’s office. McAllister then advised respondent for the first time of his Miranda rights, reading from a standard card. Respondent indicated he understood his rights, and, having these rights in mind, wished to speak with the officers. Elstad gave a full statement, explaining that he had known that the Gross family was out of town and had been paid to lead several acquaintances to the Gross residence and show them how to gain entry through a defective sliding glass door. The statement was typed, reviewed by respondent, read back to him for correction, initialed and signed by Elstad and both officers. As an afterthought, Elstad added and initialed the sentence, “After leaving the house Robby & I went back to [the] van & Robby handed *302me a small bag of grass.” App. 42. Respondent concedes that the officers made no threats or promises either at his residence or at the Sheriff’s office.

Respondent was charged with first-degree burglary. He was represented at trial by retained counsel. Elstad waived his right to a jury, and his case was tried by a Circuit Court Judge. Respondent moved at once to suppress his oral statement and signed confession. He contended that the statement he made in response to questioning at his house “let the cat out of the bag,” citing United States v. Bayer, 331 U. S. 532 (1947), and tainted the subsequent confession as “fruit of the poisonous tree,” citing Wong Sun v. United States, 371 U. S. 471 (1963). The judge ruled that the statement, “I was there,” had to be excluded because the defendant had not been advised of his Miranda rights. The written confession taken after Elstad’s arrival at the Sheriff’s office, however, was admitted in evidence. The court found:

“[H]is written statement was given freely, voluntarily and knowingly by the defendant after he had waived his right to remain silent and have counsel present which waiver was evidenced by the card which the defendant had signed. [It] was not tainted in any way by the previous brief statement between the defendant and the Sheriff’s Deputies that had arrested him.” App. 45.

Elstad was found guilty of burglary in the first degree. He received a 5-year sentence and was ordered to pay $18,000 in restitution.

Following his conviction, respondent appealed to the Oregon Court of Appeals, relying on Wong Sun and Bayer. The State conceded that Elstad had been in custody when he made his statement, “I was there,” and accordingly agreed that this statement was inadmissible as having been given without the prescribed Miranda warnings. But the State maintained that any conceivable “taint” had been dissipated prior to the respondent’s written confession by McAllister’s careful administration of the requisite warnings. The Court *303of Appeals reversed respondent’s conviction, identifying the crucial constitutional inquiry as “whether there was a sufficient break in the stream of events between [the] inadmissible statement and the written confession to insulate the latter statement from the effect of what went before.” 61 Ore. App. 673, 676, 658 P. 2d 552, 554 (1983). The Oregon court concluded:

“Regardless of the absence of actual compulsion, the coercive impact of the unconstitutionally obtained statement remains, because in a defendant’s mind it has sealed his fate. It is this impact that must be dissipated in order to make a subsequent confession admissible. In determining whether it has been dissipated, lapse of time, and change of place from the original surroundings are the most important considerations.” Id., at 677, 658 P. 2d, at 554.

Because of the brief period separating the two incidents, the “cat was sufficiently out of the bag to exert a coercive impact on [respondent’s] later admissions.” Id., at 678, 658 P. 2d, at 555.

The State of Oregon petitioned the Oregon Supreme Court for review, and review was declined. This Court granted certiorari to consider the question whether the Self-Incrimination Clause of the Fifth Amendment requires the suppression of a confession, made after proper Miranda warnings and a valid waiver of rights, solely because the police had obtained an earlier voluntary but unwarned admission from the defendant.

II

The arguments advanced in favor of; suppression of respondent’s written confession rely heavily on metaphor. One metaphor, familiar from the Fourth Amendment context, would require that respondent’s confession, regardless of its integrity, voluntariness, and probative value, be suppressed as the “tainted fruit of the poisonous tree” of the Miranda, violation. A second metaphor questions whether a *304confession can be truly voluntary once the “cat is out of the bag.” Taken out of context, each of these metaphors can be misleading. They should not be used to obscure fundamental differences between the role of the Fourth Amendment exclusionary rule and the function of Miranda in guarding against the prosecutorial use of compelled statements as prohibited by the Fifth Amendment. The Oregon court assumed and respondent here contends that a failure to administer Miranda warnings necessarily breeds the same consequences as police infringement of a constitutional right, so that evidence uncovered following an unwarned statement must be suppressed as “fruit of the poisonous tree.” We believe this view misconstrues the nature of the protections afforded by Miranda warnings and therefore misreads the consequences of police failure to supply them.

A

Prior to Miranda, the admissibility of an accused’s in-custody statements was judged solely by whether they were “voluntary” within the meaning of the Due Process Clause. See, e. g., Haynes v. Washington, 373 U. S. 503 (1963); Chambers v. Florida, 309 U. S. 227 (1940). If a suspect’s statements had been obtained by “techniques and methods offensive to due process,” Haynes v. Washington, 373 U. S., at 515, or under circumstances in which the suspect clearly had no opportunity to exercise “a free and unconstrained will,” id., at 514, the statements would not be admitted. The Court in Miranda required suppression of many statements that would have been admissible under traditional due process analysis by presuming that statements made while in custody and without adequate warnings were protected by the Fifth Amendment. The Fifth Amendment, of course, is not concerned with nontestimonial evidence. See Schmerber v. California, 384 U. S. 757, 764 (1966) (defendant may be compelled to supply blood samples). Nor is it concerned *305with moral and psychological pressures to confess emanating from sources other than official coercion. See, e. g., California v. Beheler, 463 U. S. 1121, 1125, and n. 3 (1983) (per curiam); Rhode Island v. Innis, 446 U. S. 291, 303, and n. 10 (1980); Oregon v. Mathiason, 429 U. S. 492, 495-496 (1977). Voluntary statements “remain a proper element in law enforcement.” Miranda v. Arizona, 384 U. S., at 478. “Indeed, far from being prohibited by the Constitution, admissions of guilt by wrongdoers, if not coerced, are inherently desirable. . . . Absent some officially coerced self-accusation, the Fifth Amendment privilege is not violated by even the most damning admissions.” United States v. Washington, 431 U. S. 181, 187 (1977). As the Court noted last Term in New York v. Quarles, 467 U. S. 649, 654 (1984) (footnote omitted):

“The Miranda Court, however, presumed that interrogation in certain custodial circumstances is inherently coercive and . . . that statements made under those circumstances are inadmissible unless the suspect is specifically informed of his Miranda rights and freely decides to forgo those rights. The prophylactic Miranda warnings therefore are ‘not themselves rights protected by the Constitution but [are] instead measures to insure that the right against compulsory self-incrimination [is] protected.’ Michigan v. Tucker, 417 U. S. 433, 444 (1974); see Edwards v. Arizona, 451 U. S. 477, 492 (1981) (Powell, J., concurring). Requiring Miranda warnings before custodial interrogation provides ‘practical reinforcement’ for the Fifth Amendment right.”

Respondent’s contention that his confession was tainted by the earlier failure of the police to provide Miranda warnings and must be excluded as “fruit of the poisonous tree” assumes the existence of a constitutional violation. This figure of speech is drawn from Wong Sun v. United States, 371 U. S. 471 (1963), in which the Court held that evidence and wit*306nesses discovered as a result of a search in violation of the Fourth Amendment must be excluded from evidence. The Wong Sun doctrine applies as well when the fruit of the Fourth Amendment violation is a confession. It is settled law that “a confession obtained through custodial interrogation after an illegal arrest should be excluded unless intervening events break the causal connection between the illegal arrest and the confession so that the confession is ‘sufficiently an act of free will to purge the primary taint.’” Taylor v. Alabama, 457 U. S. 687, 690 (1982) (quoting Brown v. Illinois, 422 U. S. 590, 602 (1975)).

But as we explained in Quarles and Tucker, a procedural Miranda violation differs in significant respects from violations of the Fourth Amendment, which have traditionally mandated a broad application of the “fruits” doctrine. The purpose of the Fourth Amendment exclusionary rule is to deter unreasonable searches, no matter how probative their fruits. Dunaway v. New York, 442 U. S. 200, 216-217 (1979); Brown v. Illinois, 422 U. S., at 600-602. “The exclusionary rule, . . . when utilized to effectuate the Fourth Amendment, serves interests and policies that are distinct from those it serves under the Fifth.” Id., at 601. Where a Fourth Amendment violation “taints” the confession, a finding of voluntariness for the purposes of the Fifth Amendment is merely a threshold requirement in determining whether the confession may be admitted in evidence. Taylor v. Alabama, supra, at 690. Beyond this, the prosecution must show a sufficient break in events to undermine the inference that the confession was caused by the Fourth Amendment violation.

The Miranda exclusionary rule, however, serves the Fifth Amendment and sweeps more broadly than the Fifth Amendment itself. It may be triggered even in the absence of a Fif th Amendment violation.1 The Fif th Amendment prohib*307its use by the prosecution in its case in chief only of compelled testimony. Failure to administer Miranda warnings creates a presumption of compulsion. Consequently, unwarned statements that are otherwise voluntary within the meaning of the Fifth Amendment must nevertheless be excluded from evidence under Miranda. Thus, in the individual case, Miranda’s preventive medicine provides a remedy even to the defendant who has suffered no identifiable constitutional harm. See New York v. Quarles, supra, at 654; Michigan v. Tucker, 417 U. S. 433, 444 (1974).

But the Miranda presumption, though irrebuttable for purposes of the prosecution’s case in chief, does not require that the statements and their fruits be discarded as inherently tainted. Despite the fact that patently voluntary statements taken in violation of Miranda must be excluded from the prosecution’s case, the presumption of coercion does not bar their use for impeachment purposes on cross-examination. Harris v. New York, 401 U. S. 222 (1971). The Court in Harris rejected as an “extravagant extension of the Constitution,” the theory that a defendant who had confessed under circumstances that made the confession inadmissible, could thereby enjoy the freedom to “deny every fact disclosed or discovered as a ‘fruit’ of his confession, free from confrontation with his prior statements” and that the voluntariness of his confession would be totally irrelevant. Id., at 225, and n. 2. Where an unwarned statement is preserved for use in situations that fall outside the sweep of the Miranda presumption, “the primary criterion of admissibility *308[remains] the ‘old’ due process voluntariness test.” Schul-hofer, Confessions and the Court, 79 Mich. L. Rev. 865, 877 (1981).

In Michigan v. Tucker, supra, the Court was asked to extend the Wong Sun fruits doctrine to suppress the testimony of a witness for the prosecution whose identity was discovered as the result of a statement taken from the accused without benefit of full Miranda warnings. As in respondent’s case, the breach of the Miranda procedures in Tucker involved no actual compulsion. The Court concluded that the unwarned questioning “did not abridge respondent’s constitutional privilege . . . but departed only from the prophylactic standards later laid down by this Court in Miranda to safeguard that privilege.” 417 U. S., at 446. Since there was no actual infringement of the suspect’s constitutional rights, the case was not controlled by the doctrine expressed in Wong Sun that fruits of a constitutional violation must be suppressed. In deciding “how sweeping the judicially imposed consequences” of a failure to administer Miranda warnings should be, 417 U. S., at 445, the Tucker Court noted that neither the general goal of deterring improper police conduct nor the Fifth Amendment goal of assuring trustworthy evidence would be served by suppression of the witness’ testimony. The unwarned confession must, of course, be suppressed, but the Court ruled that introduction of the third-party witness’ testimony did not violate Tucker’s Fifth Amendment rights.

We believe that this reasoning applies with equal force when the alleged “fruit” of a noncoercive Miranda violation is neither a witness nor an article of evidence but the accused’s own voluntary testimony. As in Tucker, the absence of any coercion or improper tactics undercuts the twin rationales— trustworthiness and deterrence — for a broader rule. Once warned, the suspect is free to exercise his own volition in deciding whether or not to make a statement to the authorities. The Court has often noted: “‘[A] living witness is not to be *309mechanically equated with the proffer of inanimate eviden-tiary objects illegally seized. . . . [T]he living "witness is an individual human personality whose attributes of will, perception, memory and volition interact to determine what testimony he wall give.”’ United States v. Ceccolini, 435 U. S. 268, 277 (1978) (emphasis added) (quoting from Smith v. United States, 117 U. S. App. D. C. 1, 3-4, 324 F. 2d 879, 881-882 (1963) (Burger, J.) (footnotes omitted), cert. denied, 377 U. S. 954 (1964)).

Because Miranda warnings may inhibit persons from giving information, this Court has determined that they need be administered only after the person is taken into “custody” or his freedom has otherwise been significantly restrained. Miranda v. Arizona, 384 U. S., at 478. Unfortunately, the task of defining “custody” is a slippery one, and “policemen investigating serious crimes [cannot realistically be expected to] make no errors whatsoever.” Michigan v. Tucker, supra, at 446. If errors are made by law enforcement officers in administering the prophylactic Miranda procedures, they should not breed the same irremediable consequences as police infringement of the Fifth Amendment itself. It is an unwarranted extension of Miranda to hold that a simple failure to administer the warnings, unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will, so taints the investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period. Though Miranda requires that the unwarned admission must be suppressed, the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.

B

The Oregon court, however, believed that the unwarned remark compromised the voluntariness of respondent’s later confession. It was the court’s view that the prior answer *310and not the unwarned questioning impaired respondent’s ability to give a valid waiver and that only lapse of time and change of place could dissipate what it termed the “coercive impact” of the inadmissible statement. When a prior statement is actually coerced, the time that passes between confessions, the change in place of interrogations, and the change in identity of the interrogators all bear on whether that coercion has carried over into the second confession. See Westover v. United States, decided together with Miranda v. Arizona, 384 U. S., at 494; Clewis v. Texas, 386 U. S. 707 (1967). The failure of police to administer Miranda warnings does not mean that the statements received have actually been coerced, but only that courts will presume the privilege against compulsory self-incrimination has not been intelligently exercised. See New York v. Quarles, 467 U. S., at 654, and n. 5; Miranda v. Arizona, supra, at 457. Of the courts that have considered whether a properly warned confession must be suppressed because it was preceded by an unwarned but clearly voluntary admission, the majority have explicitly or implicitly recognized that Westover's requirement of a break in the stream of events is inapposite.2 In these circumstances, a careful and thorough *311administration of Miranda warnings serves to cure the condition that rendered the unwarned statement inadmissible. The warning conveys the relevant information and thereafter the suspect’s choice whether to exercise his privilege to remain silent should ordinarily be viewed as an “act of free will.” Wong Sun v. United States, 371 U. S., at 486.

The Oregon court nevertheless identified a subtle form of lingering compulsion, the psychological impact of the suspect’s conviction that he has let the cat out of the bag and, in so doing, has sealed his own fate. But endowing the psychological effects of voluntary unwarned admissions with constitutional implications would, practically speaking, disable the police from obtaining the suspect’s informed cooperation even when the official coercion proscribed by the Fifth Amendment played no part in either his warned or unwarned confessions. As the Court remarked in Bayer:

“[AJfter an accused has once let the cat out of the bag by confessing, no matter what the inducement, he is never thereafter free of the psychological and practical disadvantages of having confessed. He can never get the cat back in the bag. The secret is out for good. In such a sense, a later confession may always be looked upon as fruit of the first. But this Court has never gone so far as to hold that making a confession under circumstances which preclude its use, perpetually disables the confessor from making a usable one after those conditions have been removed.” 331 U. S., at 540-541.

Even in such extreme cases as Lyons v. Oklahoma, 322 U. S. 596 (1944), in which police forced a full confession from the accused through unconscionable methods of interrogation, the Court has assumed that the coercive effect of the confes*312sion could, with time, be dissipated. See also Westover v. United States, supra, at 496.

This Court has never held that the psychological impact of voluntary disclosure of a guilty secret qualifies as state compulsion or compromises the voluntariness of a subsequent informed waiver. The Oregon court, by adopting this expansive view of Fifth Amendment compulsion, effectively immunizes a suspect who responds to pre-Miranda warning questions from the consequences of his subsequent informed waiver of the privilege of remaining silent. See 61 Ore. App., at 679, 658 P. 2d, at 555 (Gillette, P. J., concurring). This immunity comes at a high cost to legitimate law enforcement activity, while adding little desirable protection to. the individual’s interest in not being compelled to testify against himself. Cf. Michigan v. Mosley, 423 U. S. 96, 107-111 (1975) (White, J., concurring in result). When neither the initial nor the subsequent admission is coerced, little, justification exists for permitting the highly probative evidence of a voluntary confession to be irretrievably lost to the factfinder.

There is a vast difference between the direct consequences flowing from coercion of a confession by physical violence or other deliberate means calculated to break the suspect’s will and the uncertain consequences of disclosure of a “guilty secret” freely given in response to an unwarned but non-coercive question, as in this case. Justice Brennan’s contention that it is impossible to perceive any causal distinction between this case and one involving a confession that is coerced by torture is wholly unpersuasive.3 Certainly, in *313respondent’s case, the causal connection between any psychological disadvantage created by his admission and his ultimate decision to cooperate is speculative and attenuated at

*314Though belated, the reading of respondent’s rights was undeniably complete. McAllister testified that he read the Miranda warnings aloud from a printed card and recorded best. It is difficult to tell with certainty what motivates a suspect to speak. A suspect’s confession may be traced to factors as disparate as “a prearrest event such as a visit with a minister,” Dunaway v. New York, 442 U. S., at 220 (Stevens, J., concurring), or an intervening event such as the exchange of words respondent had with his father. We must conclude that, absent deliberately coercive or improper tactics in obtaining the initial statement, the mere fact that a suspect has made an unwarned admission does not warrant a presumption of compulsion. A subsequent administration of Miranda warnings to a suspect who has given a voluntary but unwarned statement ordinarily should suffice to remove the conditions that precluded admission of the earlier statement. In such circumstances, the finder of fact may reasonably conclude that the suspect made a rational and intelligent choice whether to waive or invoke his rights.

I — I *315Elstad’s responses.4 There is no question that respondent knowingly and voluntarily waived his right to remain silent before he described his participation in the burglary. It is also beyond dispute that respondent’s earlier remark was voluntary, within the meaning of the Fifth Amendment. Neither the environment nor the manner of either “interrogation” was coercive. The initial conversation took place at midday, in the living room area of respondent’s own home, with his mother in the kitchen area, a few steps away. Although in retrospect the officers testified that respondent was then in custody, at the time he made his statement he had not been informed that he was under arrest. The arresting officers’ testimony indicates that the brief stop in the living room before proceeding to the station house was not to interrogate the suspect but to notify his mother of the reason for his arrest. App. 9-10.

The State has conceded the issue of custody and thus we must assume that Burke breached Miranda procedures in failing to administer Miranda warnings before initiating the discussion in the living room. This breach may have been the result of confusion as to whether the brief exchange qualified as “custodial interrogation” or it may simply have reflected Burke’s reluctance to initiate an alarming police *316procedure before McAllister had spoken with respondent’s mother. Whatever the reason for Burke’s oversight, the incident had none of the earmarks of coercion. See Rawlings v. Kentucky, 448 U. S. 98, 109-110 (1980). Nor did the officers exploit the unwarned admission to pressure respondent into waiving his right to remain silent.

Respondent, however, has argued that he was unable to give a fully informed waiver of his rights because he was unaware that his prior statement could not be used against him. Respondent suggests that Officer McAllister, to cure this deficiency, should have added an additional warning to those given him at the Sheriff’s office. Such a requirement is neither practicable nor constitutionally necessary. In many cases, a breach of Miranda procedures may not be identified as such until long after full Miranda warnings are administered and a valid confession obtained. See, e. g., United States v. Bowler, 561 F. 2d 1323, 1324-1325 (CA9 1977) (certain statements ruled inadmissible by trial court); United States v. Toral, 536 F. 2d 893, 896 (CA9 1976); United States v. Knight, 395 F. 2d 971, 974-975 (CA2 1968) (custody unclear). The standard Miranda warnings explicitly inform the suspect of his right to consult a lawyer before speaking. Police officers are ill-equipped to pinch-hit for counsel, construing the murky and difficult questions of when “custody” begins or whether a given unwarned statement will ultimately be held admissible. See Tanner v. Vincent, 541 F. 2d 932, 936 (CA2 1976), cert. denied, 429 U. S. 1065 (1977).

This Court has never embraced the theory that a defendant’s ignorance of the full consequences of his decisions vitiates their voluntariness. See California v. Beheler, 463 U. S., at 1125-1126, n. 3; McMann v. Richardson, 397 U. S. 759, 769 (1970). If the prosecution has actually violated the defendant’s Fifth Amendment rights by introducing an inadmissible confession at trial, compelling the defendant to testify in rebuttal, the rule announced in Harrison v. United States, 392 U. S. 219 (1968), precludes use of that testimony *317on retrial. “Having ‘released the spring’ by using the petitioner’s unlawfully obtained confessions against him, the Government must show that its illegal action did not induce his testimony.” Id., at 224-225. But the Court has refused to find that a defendant who confesses, after being falsely told that his codefendant has turned State’s evidence, does so involuntarily. Frazier v. Cupp, 394 U. S. 731, 739 (1969). The Court has also rejected the argument that a defendant’s ignorance that a prior coerced confession could not be admitted in evidence compromised the voluntariness of his guilty plea. McMann v. Richardson, supra, at 769. Likewise, in California v. Beheler, supra, the Court declined to accept defendant’s contention that, because he was unaware of the potential adverse consequences of statements he made to the police, his participation in the interview was involuntary. Thus we have not held that the sine qua non for a knowing and voluntary waiver of the right to remain silent is a full and complete appreciation of all of the consequences flowing from the nature and the quality of the evidence in the case.

J — I <1

When police ask questions of a suspect in custody without administering the required warnings, Miranda dictates that the answers received be presumed compelled and that they be excluded from evidence at trial in the State’s case in chief. The Court has carefully adhered to this principle, permitting a narrow exception only where pressing public safety concerns demanded. See New York v. Quarles, 467 U. S., at 655-656. The Court today in no way retreats from the bright-line rule of Miranda. We do not imply that good faith excuses, a failure to administer Miranda warnings; nor do we condone inherently coercive police tactics or methods offensive to due process that render the initial admission involuntary and undermine the suspect’s will to invoke his rights once they are read to him. A handful of courts have, however, applied our precedents relating to confessions ob*318tained under coercive circumstances to situations involving wholly voluntary admissions, requiring a passage of time or break in events before a second, fully warned statement can be deemed voluntary. Far from establishing a rigid rule, we direct courts to avoid one; there is no warrant for presuming coercive effect where the suspect’s initial inculpatory statement, though technically in violation of Miranda, was voluntary.5 The relevant inquiry is whether, in fact, the second statement was also voluntarily made. As in any such inquiry, the finder of fact must examine the surrounding circumstances and the entire course of police conduct with respect to the suspect in evaluating the voluntariness of his statements. The fact that a suspect chooses to speak after being informed of his rights is, of course, highly probative. We find that the dictates of Miranda and the goals of the Fifth Amendment proscription against use of compelled testimony are fully satisfied in the circumstances of this case by barring use of the unwarned statement in the case in chief. No further purpose is served by imputing “taint” to subsequent statements obtained pursuant to a voluntary and knowing waiver. We hold today that a suspect who has once responded to unwarned yet uncoercive questioning is not thereby disabled from waiving his rights and confessing after he has been given the requisite Miranda warnings.

The judgment of the Court of Appeals of Oregon is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

Justice Brennan,

with whom Justice Marshall joins,

dissenting.

The Self-Incrimination Clause of the Fifth Amendment guarantees every individual that, if taken into official cus*319tody, he shall be informed of important constitutional rights and be given the opportunity knowingly and voluntarily to waive those rights before being interrogated about suspected wrongdoing. Miranda v. Arizona, 384 U. S. 436 (1966).1 This guarantee embodies our society’s conviction that “no system of criminal justice can, or should, survive if it comes to depend for its continued effectiveness on the citizens’ abdication through unawareness of their constitutional rights.” Escobedo v. Illinois, 378 U. S. 478, 490 (1964).

Even while purporting to reaffirm these constitutional guarantees, the Court has engaged of late in a studied campaign to strip the Miranda decision piecemeal and to undermine the rights Miranda sought to secure. Today’s decision not only extends this effort a further step, but delivers a potentially crippling blow to Miranda and the ability of courts to safeguard the rights of persons accused of crime. For at least with respect to successive confessions, the Court today appears to strip remedies for Miranda violations of the “fruit of the poisonous tree” doctrine prohibiting the use of evidence presumptively derived from official illegality.2

Two major premises undergird the Court’s decision. The Court rejects as nothing more than “speculative” the long-recognized presumption that an illegally extracted confession causes the accused to confess again out of the mistaken belief that he already has sealed his fate, and it condemns as “ ‘extravagant’ ” the requirement that the prosecution affirmatively rebut the presumption before the subsequent confes*320sion may be admitted. Ante, at 307, 313. The Court instead adopts a new rule that, so long as the accused is given the usual Miranda warnings before further interrogation, the taint of a previous confession obtained in violation oí Miranda “ordinarily” must be viewed as automatically dissipated. Ante, at 311.

In the alternative, the Court asserts that neither the Fifth Amendment itself nor the judicial policy of deterring illegal police conduct requires the suppression of the “fruits” of a confession obtained in violation of Miranda, reasoning that to do otherwise would interfere with “legitimate law enforcement activity.” Ante, at 312. As the Court surely understands, however, “[t]o forbid the direct use of methods . . . but to put no curb on their full indirect use would only invite the very methods deemed ‘inconsistent with ethical standards and destructive of personal liberty/” Nardone v. United States, 308 U. S. 338, 340 (1939). If violations of constitutional rights may not be remedied through the well-established rules respecting derivative evidence, as the Court has held today, there is a critical danger that the rights will be rendered nothing more than a mere “form of words.” Silverthorne Lumber Co. v. United States, 251 U. S. 385, 392 (1920).

The Court’s decision says much about the way the Court currently goes about implementing its agenda. In imposing its new rule, for example, the Court mischaracterizes our precedents, obfuscates the central issues, and altogether ignores the practical realities of custodial interrogation that have led nearly every lower court to reject its simplistic reasoning. Moreover, the Court adopts startling and unprecedented methods of construing constitutional guarantees. Finally, the Court reaches out once again to address issues not before us. For example, although the State of Oregon has conceded that the arresting officers broke the law in this case, the Court goes out of its way to suggest that they may have been objectively justified in doing so.

*321Today’s decision, in short, threatens disastrous consequences far beyond the outcome in this case. As the Court has not seen fit to provide a full explanation for this result, I believe it essential to consider in detail the premises, reasoning, and implications of the Court’s opinion.

J — H

The threshold question is this: What effect should an admission or confession of guilt obtained in violation of an accused’s Miranda rights be presumed to have upon the vol-untariness of subsequent confessions that are preceded by Miranda warnings? Relying on the “cat out of the bag” analysis of United States v. Bayer, 331 U. S. 532, 540-541 (1947), the Oregon Court of Appeals held that the first confession presumptively taints subsequent confessions in such circumstances. 61 Ore. App. 673, 676, 658 P. 2d 552, 554 (1983). On the specific facts of this case, the court below found that the prosecution had not rebutted this presumption. Rather, given the temporal proximity of Elstad’s second confession to his first and the absence of any significant intervening circumstances, the court correctly concluded that there had not been “a sufficient break in the stream of events between [the] inadmissible statement and the written confession to insulate the latter statement from the effect of what went before.” Ibid.

If this Court’s reversal of the judgment below reflected mere disagreement with the Oregon court’s application of the “cat out of the bag” presumption to the particular facts of this case, the outcome, while clearly erroneous, would be of little lasting consequence. But the Court rejects the “cat out of the bag” presumption entirely and instead adopts a new rule presuming that “ordinarily” there is no causal connection between a confession extracted in violation of Miranda and a subsequent confession preceded by the usual Miranda warnings. Ante, at 311, 314. The Court suggests that it is merely following settled lower-court practice in adopting this *322rule and that the analysis followed by the Oregon Court of Appeals was aberrant. This is simply not so. Most federal courts have rejected the Court’s approach and instead held that (1) there is a rebuttable presumption that a confession obtained in violation of Miranda taints subsequent confessions, and (2) the taint cannot be dissipated solely by giving Miranda warnings.3 Moreover, those few federal courts that have suggested approaches similar to the Court’s have subsequently qualified their positions.4 Even more significant is the case among state courts. Although a handful have adopted the Court’s approach,5 the overwhelming ma*323jority of state courts that have considered the issue have concluded that subsequent confessions are presumptively tainted by a first confession taken in violation of Miranda and that Miranda warnings alone cannot dissipate the taint.6

*324The Court today sweeps aside this common-sense approach as “speculative” reasoning, adopting instead a rule that “the psychological impact of voluntary disclosure of a guilty secret” neither “qualifies as state compulsion” nor “compromises the voluntariness” of subsequent confessions. Ante, at 312, 313 (emphasis added). So long as a suspect receives the usual Miranda warnings before further interrogation, the Court reasons, the fact that he “is free to exercise his own volition in deciding whether or not to make” further confessions “ordinarily” is a sufficient “cure” and serves to break any causal connection between the illegal confession and subsequent statements. Ante, at 308, 311.

The Court’s marble-palace psychoanalysis is tidy, but it flies in the face of our own precedents, demonstrates a startling unawareness of the realities of police interrogation, and is completely out of tune with the experience of state and federal courts over the last 20 years. Perhaps the Court has grasped some psychological truth that has eluded persons far more experienced in these matters; if so, the Court owes an explanation of how so many could have been so wrong for so many years.

A

(1)

This Court has had long experience with the problem of confessions obtained after an earlier confession has been *325illegally secured. Subsequent confessions in these circumstances are not per se inadmissible, but the prosecution must demonstrate facts “sufficient to insulate the [subsequent] statement from the effect of all that went before.” Clewis v. Texas, 386 U. S. 707, 710 (1967). If the accused’s subsequent confession was merely the culmination of “one continuous process,” or if the first confession was merely “filled in and perfected by additional statements given in rapid succession,” the subsequent confession is inadmissible even though it was not obtained through the same illegal means as the first. Leyra v. Denno, 347 U. S. 556, 561 (1954); see also Westover v. United States, decided together with Miranda v. Arizona, 384 U. S. 436, 494-496 (1966). The question in each case is whether the accused’s will was “overborne at the time he confessed,” and the prosecution must demonstrate that the second confession “was an act independent of the [earlier] confession.” Reck v. Pate, 367 U. S. 433, 440, 444 (1961).

One of the factors that can vitiate the voluntariness of a subsequent confession is the hopeless feeling of an accused that he has nothing to lose by repeating his confession, even where the circumstances that rendered his first confession illegal have been removed. As the Court observed in United States v. Bayer, 331 U. S., at 540:

“[A]fter an accused has once let the cat out of the bag by confessing, no matter what the inducement, he is never thereafter free of the psychological and practical disadvantages of having confessed. He can never get the cat back in the bag. The secret is out for good. In such a sense, a later confession always may be looked upon as a fruit of the first.”

The Court today decries the “irremediable consequences” of this reasoning, ante, at 309, but it has always been clear that even after “let[ting] the cat out of the bag” the accused is not “perpetually disable[d]” from giving an admissible subsequent confession. United States v. Bayer, supra, at 541. *326Rather, we have held that subsequent confessions in such circumstances may be admitted if the prosecution demonstrates that, “[cjonsidering the ‘totality of the circumstances,’ ” there was a “‘break in the stream of events . . . sufficient to insulate’ ” the subsequent confession from the damning impact of the first. Darwin v. Connecticut, 391 U. S. 346, 349 (1968) (citations omitted). Although we have thus rejected a 'per se rule forbidding the introduction of subsequent statements in these circumstances, we have emphasized that the psychological impact of admissions and confessions of criminal guilt nevertheless can have a decisive impact in undermining the voluntariness of a suspect’s responses to continued police interrogation and must be accounted for in determining their admissibility. As Justice Harlan explained in his separate Darwin opinion:

“A principal reason why a suspect might make a second or third confession is simply that, having already confessed once or twice, he might think he has little to lose by repetition. If a first confession is not shown to be voluntary, I do not think a later confession that is merely a direct product of the earlier one should be held to be voluntary. It would be neither conducive to good police work, nor fair to a suspect, to allow the erroneous impression that he has nothing to lose to play the major role in a defendant’s decision to speak a second or third time.

“In consequence, when the prosecution seeks to use a confession uttered after an earlier one not found to be voluntary, it has . . . the burden of proving not only that the later confession was not itself the product of improper threats or promises or coercive conditions, but also that it was not directly produced by the existence of the earlier confession.” Id., at 350-351 (concurring in part and dissenting in part).

See also Brown v. Illinois, 422 U. S. 590, 605, n. 12 (1975) (“The fact that Brown had made one statement, believed by *327him to be admissible, . . . bolstered the pressures for him to give the second, or at least vitiated any incentive on his part to avoid self-incrimination”); Beecher v. Alabama, 389 U. S. 35, 36, n. 2 (1967) (per curiam) (existence of earlier illegal confession “is of course vitally relevant to the voluntariness of petitioner’s later statements”).7

*328(2)

Our precedents did not develop in a vacuum. They reflect an understanding of the realities of police interrogation and the everyday experience of lower courts. Expert interrogators, far from dismissing a first admission or confession as creating merely a “speculative and attenuated” disadvantage for a suspect, ante, at 313, understand that such revelations frequently lead directly to a full confession. Standard interrogation manuals advise that “[t]he securing of the first admission is the biggest stumbling block . . . .” A. Aubry & R. Caputo, Criminal Interrogation 290 (3d ed. 1980). If this first admission can be obtained, “there is every reason to expect that the first admission will lead to others, and eventually to the full confession.” Ibid.

“For some psychological reason which does not have to concern us at this point ‘the dam finally breaks as a result of the first leak’ with regards to the tough subject. . . . Any structure is only as strong as its weakest component, and total collapse can be anticipated when the weakest part first begins to sag.” Id., at 291.

Interrogators describe the point of the first admission as the “breakthrough” and the “beachhead,” R. Royal & S. Schutt, The Gentle Art of Interviewing and Interrogation: A Professional Manual and Guide 143 (1976), which once obtained will give them enormous “tactical advantages,” F. Inbau & J. Reid, Criminal Interrogation and Confessions 82 (2d ed. 1967). See also W. Dienstein, Technics for the Crime Investigator 117 (2d ed. 1974). Thus “[t]he securing of incriminating admissions might well be considered as the beginning of the final stages in crumbling the defenses of the suspect,” and the process of obtaining such admissions is described as “the spadework required to motivate the subject into making the full confession.” Aubry & Caputo, supra, at 31, 203.

*329“Once the initial admission has been made, further inducement in the form of skillfully applied interrogation techniques will motivate the suspect into making the confession.” Id., at 26; see also id., at 33 (initial admissions are “capitalized upon by the interrogator in securing the eventual confession”). Some of these “skillfully applied” techniques involve direct confrontation of the suspect with the earlier admission, but many of the techniques are more discreet and create leverage without the need of expressly discussing the earlier admission. These techniques are all aimed at reinforcing in the suspect’s mind that, as one manual describes it, “ ‘you’re wasting your own time, and you’re wasting my time, you’re guilty and you know it, I know it, what’s more, you know that I know it.’” Id., at 234.8

The practical experience of state and federal courts confirms the experts’ understanding. From this experience, lower courts have concluded that a first confession obtained without proper Miranda warnings, far from creating merely some “speculative and attenuated” disadvantage for the accused, ante, at 313, frequently enables the authorities to obtain subsequent confessions on a “silver platter.” Cagle v. State, 45 Ala. App. 3, 4, 221 So. 2d 119,120, cert. denied, 284 Ala. 727, 221 So. 2d 121 (1969).

One police practice that courts have frequently encountered involves the withholding of Miranda warnings until the end of an interrogation session. Specifically, the police *330escort a suspect into a room, sit him down and, without explaining his Fifth Amendment rights or obtaining a knowing and voluntary waiver of those rights, interrogate him about his suspected criminal activity. If the police obtain a confession, it is then typed up, the police hand the suspect a pen for his signature, and — -just before he signs — the police advise him of his Miranda rights and ask him to proceed. Alternatively, the police may call a stenographer in after they have obtained the confession, advise the suspect for the first time of his Miranda rights, and ask him to repeat what he has just told them. In such circumstances, the process of giving Miranda warnings and obtaining the final confession is “ ‘merely a formalizing, a setting down almost as a scrivener does, [of] what ha[s] already taken [place].’ ” People v. Raddatz, 91 Ill. App. 2d 425, 430, 235 N. E. 2d 353, 356 (1968) (quoting trial court). In such situations, where “it was all over except for reading aloud and explaining the written waiver of the Miranda safeguards,” courts have time and again concluded that “[t]he giving of the Miranda warnings before reducing the product of the day’s work to written form could not undo what had been done or make legal what was illegal.” People v. Bodner, 75 App. Div. 2d 440, 448, 430 N. Y. S. 2d 433, 438 (1980).9

There are numerous variations on this theme. Police may obtain a confession in violation of Miranda and then take a break for lunch or go home for the evening. When questioning is resumed, this time preceded by Miranda warnings, the suspect is asked to “clarify” the earlier illegal confession and to provide additional information.10 Or he is led by one of *331the interrogators into another room, introduced to another official, and asked to repeat his story. The new officer then gives the Miranda warnings and asks the suspect to proceed.11 Alternatively, the suspect might be questioned by arresting officers “in the field” and without Miranda warnings, as was young Elstad in the instant case. After making incriminating admissions or a confession, the suspect is then brought into the station house and either questioned by the same officers again or asked to repeat his earlier statements to another officer.12

The variations of this practice are numerous, but the underlying problem is always the same: after hearing the witness testimony and considering the practical realities, courts have confirmed the time-honored wisdom of presuming that a first illegal confession “taints” subsequent confessions, and permitting such subsequent confessions to be admitted at trial only if the prosecution convincingly rebuts the presumption. They have discovered that frequently, “[h]aving once confessed [the accused] was ready to confess some more.” State v. Lekas, 201 Kan. 579, 587-588, 442 P. 2d 11, 19 (1968). For all practical purposes, the prewaming and postwarning questioning are often but stages of one overall interrogation. Whether or not the authorities explicitly confront the suspect with his earlier illegal admissions makes no significant difference, of course, because the suspect knows that the authorities know of his earlier statements and most frequently will believe that those statements already have sealed his fate. Thus a suspect in such circumstances is likely to conclude that “he might as well answer the questions *332put to him, since the [authorities are] already aware of the earlier answers,” United States v. Pierce, 397 F. 2d 128, 131 (CA4 1968); he will probably tell himself that “it’s O. K., I have already told them,” State v. Lekas, supra, at 582, 442 P. 2d, at 15. See also Cagle v. State, 45 Ala. App., at 4, 221 So. 2d, at 120 (“I have already give[n] the Chief... a statement, and I might as well give one to you, too”). In such circumstances, courts have found, a suspect almost invariably asks himself, “What use is a lawyer? What good is a lawyer now? What benefit can a lawyer tell me? [sic] I have already told the police everything.” People v. Raddatz, 91 Ill. App. 2d, at 430, 235 N. E. 2d, at 356.13

I would have thought that the Court, instead of dismissing the “cat out of the bag” presumption out of hand, would have accounted for these practical realities. Compare Nardone v. United States, 308 U. S., at 342 (derivative-evidence rules should be grounded on the “learning, good sense, fairness and courage” of lower-court judges). Expert interrogators and experienced lower-court judges will be startled, to say the least, to learn that the connection between multiple confessions is “speculative” and that a subsequent rendition of Miranda warnings “ordinarily” enables the accused in these circumstances to exercise his “free will” and to make “a rational and intelligent choice whether to waive or invoke his rights.” Ante, at 311, 314.

(3)

The Court’s new view about the “psychological impact” of prior illegalities also is at odds with our Fourth Amendment *333precedents. For example, it is well established that a confession secured as a proximate result of an illegal arrest must be suppressed. See, e. g., Taylor v. Alabama, 457 U. S. 687 (1982); Brown v. Illinois, 422 U. S. 590 (1975); Wong Sun v. United States, 371 U. S. 471 (1963). We have emphasized in this context that “verbal evidence which derives so immediately from an unlawful entry and an unauthorized arrest ... is no less the ‘fruit’ of official illegality than the more common tangible fruits of the unwarranted intrusion.” Wong Sun v. United States, supra, at 485.

The Court seeks to distinguish these precedents on the ground that Fourth Amendment violations require a broader exclusionary rule than do Fifth Amendment violations. Ante, at 306. I address this reasoning in Part II-B, infra. But the question immediately at issue — whether there should be a presumptive rule against finding a causal connection between successive confessions — would surely seem to be controlled by the logic of these Fourth Amendment cases. In part because of the inherent psychological pressures attendant upon an arrest, we have refused to presume that a confession following an illegal arrest is “sufficiently an act of free will to purge the primary taint of the unlawful invasion.” Wong Sun v. United States, supra, at 486. See also Brown v. Illinois, supra, at 601-603. If the Court so quickly dismisses the notion of a multiple-confession taint as nothing more than a “speculative and attenuated” disadvantage, ante, at 313, what is to prevent it in the future from deciding that, contrary to the settled understanding, the fact of a proximate illegal arrest is presumptively nothing but a “speculative and attenuated” disadvantage to a defendant who is asked to confess?

Similarly, a confession obtained as a proximate result of confronting the accused with illegally seized evidence is inadmissible as the fruit of the illegal seizure. See, e. g., Fahy v. Connecticut, 375 U. S. 85, 90-91 (1963) (remanding for determination whether admission was so induced); see generally 3 W. LaFave, Search and Seizure §11.4, pp. 638-642 *334(1978) (collecting cases). As commentators have noted, courts in finding such confessions to be tainted by the Fourth Amendment violation have emphasized that “ ‘the realization that the “cat is out of the bag” plays a significant role in encouraging the suspect to speak.’” Id., § 11.4, p. 689 (footnote omitted). By discarding the accepted “cat out of the bag” presumption in the successive-confession context, however, the Court now appears to have opened the door to applying this same simplistic reasoning to Fourth Amendment violations.14

*335B

The correct approach, administered for almost 20 years by most courts with no untoward results, is to presume that an admission or confession obtained in violation of Miranda taints a subsequent confession unless the prosecution can show that the taint is so attenuated as to justify admission of the subsequent confession. See cases cited in nn. 3, 6, swpra. Although the Court warns against the “irremediable consequences” of this presumption, ante, at 309, it is obvious that a subsequent confession, just like any other evidence that follows upon illegal police action, does not become “sacred and inaccessible.” Silverthorne Lumber Co. v. United States, 251 U. S., at 392. As with any other evidence, the inquiry is whether the subsequent confession “‘has been come at by exploitation of [the] illegality or instead by means sufficiently distinguishable to be purged of the primary taint.’” Wong Sun v. United States, 371 U. S., at 488 (citation omitted).

Until today the Court has recognized that the dissipation inquiry requires the prosecution to demonstrate that the official illegality did not taint the challenged confession, and we have rejected the simplistic view that abstract notions of “free will” are alone sufficient to dissipate the challenged taint.

“The question whether a confession is the product of a free will under Wong Sun must be answered on the facts of each case. No single fact is dispositive. The work*336ings of the human mind are too complex, and the possibilities of misconduct too diverse, to permit protection of [constitutional rights] to turn on... a talismanic test.” Brown v. Illinois, 422 U. S., at 603.

Instead, we have instructed courts to consider carefully such factors as the strength of the causal connection between the illegal action and the challenged evidence, their proximity in time and place, the presence of intervening factors, and the “purpose and flagrancy of the official misconduct.” Id., at 603-604.

The Court today shatters this sensitive inquiry and decides instead that, since individuals possess “‘will, perception, memory and volition/” a suspect’s “exercise [of] his own volition in deciding whether or not to make a [subsequent] statement to the authorities” must “ordinarily” be viewed as sufficient to dissipate the coercive influence of a prior confession obtained in violation of Miranda. Ante, at 308, 309, 311 (citation omitted). But “[w]ill, perception, memory and volition are only relevant as they provide meaningful alternatives in the causal chain, not as mystical qualities which in themselves invoke the doctrine of attenuation.” Hirtle, Inadmissible Confessions and Their Fruits: A Comment on Harrison v. United States, 60 J. Grim. L., C., & P. S. 58, 62 (1969). Thus we have always rejected, until today, the notion that “individual will” alone presumptively serves to insulate a person’s actions from the taint of earlier official illegality. See, e. g., United States v. Ceccolini, 435 U. S. 268, 274-275 (1978) (rejecting Government’s request for a rule “that the testimony of a live witness should not be excluded at trial no matter how close and proximate the connection between it” and an illegal search); Wong Sun v. United States, swpra, at 486 (confession obtained as a proximate result of an illegal arrest is not presumptively admissible as an “intervening independent act of a free will”).

Nor have we ever allowed Miranda warnings alone to serve talismanically to purge the taint of prior illegalities. In Brown v. Illinois, for example, we emphasized that *337 “Miranda warnings, alone and per se, cannot always make [a confession] sufficiently a product of free will to break . . . the causal connection between [an illegal arrest] and the confession.” 422 U. S., at 603 (emphasis in original).15 See also Taylor v. Alabama, 457 U. S., at 690-691. The reason we rejected this rule is manifest: “The Miranda warnings in no way inform a person of his Fourth Amendment rights, including his right to be released from unlawful custody following an arrest made without a warrant or without probable cause.” Brown v. Illinois, supra, at 601, n. 6.

This logic applies with even greater force to the Fifth Amendment problem of successive confessions. Where an accused believes that it is futile to resist because the authorities already have elicited an admission of guilt, the mere rendition of Miranda warnings does not convey the information most critical at that point to ensuring his informed and voluntary decision to speak again: that the earlier confession may not be admissible and thus that he need not speak out of any feeling that he already has sealed his fate. The Court therefore is flatly wrong in arguing, as it does repeatedly, that the mere provision of Miranda warnings prior to subsequent interrogation supplies the accused with “the relevant information” and ensures that a subsequent confession “ordinarily” will be the product of “a rational and intelligent choice” and “‘an act of free will.’” Ante, at 311, 314.16

*338The Court’s new approach is therefore completely at odds with established dissipation analysis. A comparison of the Court’s analysis with the factors most frequently relied on by lower courts in considering the admissibility of subsequent confessions demonstrates the practical and legal flaws of the new rule.

Advice that earlier confession may be inadmissible. The most effective means to ensure the voluntariness of an accused’s subsequent confession is to advise the accused that his earlier admissions may not be admissible and therefore that he need not speak solely out of a belief that “the cat is out of the bag.” Many courts have required such warnings in the absence of other dissipating factors,17 and this Court has not uncovered anything to suggest that this approach has not succeeded in the real world. The Court, however, believes that law enforcement authorities could never possibly understand “the murky and difficult questio[n]” of when *339 Miranda warnings must be given, and therefore that they are “ill-equipped” to make the decision whether supplementary warnings might be required. Ante, at 316.

This reasoning is unpersuasive for two reasons. First, the whole point of Miranda and its progeny has been to prescribe “bright line” rules for the authorities to follow.18 Although borderline cases will of course occasionally arise, thus militating against a per se rule requiring supplementary warnings, the experience of the lower courts demonstrates that the vast majority of confrontations implicating this question involve obvious Miranda violations. The occasional “murky and difficult” case should not preclude consideration of supplementary warnings in situations where the authorities could not possibly have acted in an objectively reasonable manner in their earlier interrogation of the accused. Second, even where the authorities are not certain that an earlier confession has been illegally obtained, courts and commentators have recognized that a supplementary warning merely advising the accused that his earlier confession may be inadmissible can dispel his belief that he has nothing to lose by repetition.19

Proximity in time and place. Courts have frequently concluded that a subsequent confession was so removed in time and place from the first that the accused most likely was able fully to exercise his independent judgment in deciding whether to speak again.20 As in the instant case, however, a *340second confession frequently follows immediately on the heels of the first and is obtained by the same officials in the same or similar coercive surroundings. In such situations, it is wholly unreasonable to assume that the mere rendition of Miranda warnings will safeguard the accused’s freedom of action.

The Court today asserts, however, that the traditional requirement that there be a “break in the stream of events” is “inapposite” in this context. Ante, at 310. Yet most lower courts that have considered the question have recognized that our decision in Westover v. United States, 384 U. S., at 494, compels the contrary conclusion.21 There the accused was questioned by local authorities for several hours and then turned over to federal officials, who only then advised him of his constitutional rights and obtained a confession. We concluded that Westover’s waiver was invalid because, from Westover’s perspective, the separate questioning amounted to but one continuous period of interrogation, “the warnings came at the end of the interrogation process,” and the giving of warnings could not dissipate the effect of *341the earlier, illegal questioning. Id., at 496.22 Thus it is clear that Miranda warnings given at the end of the interrogation process cannot dispel the illegality of what has gone before. If this is so in a situation like Westover, where the accused had not yet given a confession, how can the Court possibly conclude otherwise where the accused already has confessed and therefore feels that he has nothing to lose by “confess[ing] some more?” State v. Lekas, 201 Kan., at 588, 442 P. 2d, at 19.

Intervening factors. Some lower courts have found that because of intervening factors — such as consultation with a lawyer or family members, or an independent decision to speak — an accused’s subsequent confession could not fairly be attributed to the earlier statement taken in violation of Miranda. 23 On the other hand, where as here an accused has continuously been in custody and there is no legitimate suggestion of an intervening event sufficient to break the impact of the first confession, subsequent confessions are inadmissible.24 The Court reasons, however, that because “[a] suspect’s confession may be traced to ... an intervening event,” it “must [be] conclude[d]” that subsequent Miranda warnings presumptively enable the suspect to make “a rational and intelligent choice” whether to repeat his confession. Ante, at 314 (emphasis added). In applying the intervening-events inquiry, however, “courts must use a surgeon’s scalpel and not a meat axe.” Cf. 3 W. LaFave, Search and Seizure § 11.4, p. 624 (1978). The only proper inquiry is whether a meaningful intervening event actually occurred, not whether *342a court simply chooses to shut its eyes to human nature and the realities of custodial interrogation.

Purpose and flagrancy of the illegality. Courts have frequently taken the “purpose and flagrancy of the official misconduct” into account in considering whether the taint of illegal action was sufficiently dissipated to render a confession admissible. Brown v. Illinois, 422 U. S., at 604. In part, this inquiry has reflected conviction that particularly egregious misconduct must be deterred through particularly stern action. This factor is also important, however, because it is fair to presume that if the authorities acted flagrantly in violating the law they probably did so for ulterior motives. Thus if the authorities blatantly failed to advise an accused of his constitutional rights while interrogating him and gave him the Miranda warnings only as they handed him a typed confession for his signature, it is fair to presume that they pursued their strategy precisely to weaken his ability knowingly and voluntarily to exercise his constitutional rights.

C

Perhaps because the Court is discomfited by the radical implications of its failure to apply the settled derivative-evidence presumption to violations of Miranda, it grudgingly qualifies its sweeping pronouncements with the acknowledgment that its new presumption about so-called “ordinary” Miranda violations can be overcome by the accused. Ante, at 311, 314. Explicitly eschewing “a per se rule,” ante, at 317, the Court suggests that its approach should not be followed where the police have employed “improper tactics” or “inherently coercive methods” that are “calculated to undermine the suspect’s ability to exercise his free will.” Ante, at 308, 309, 312, n. 3; see also ante, at 312, 314, 317. The Court thus concedes that lower courts must continue to be free to “examine the surrounding circumstances and the *343entire course of police conduct with respect to the suspect in evaluating the voluntariness of his statements.” Ante, at 318.

The Court’s concessions are potentially significant, but its analysis is wholly at odds with established dissipation analysis. To begin with, the Court repeatedly suggests that a confession may be suppressed only if the police have used “improper tactics,” ante, at 308; this obscure reasoning overlooks the fact that a violation of Miranda is obviously itself an “improper tactic,” one frequently used precisely to undermine the voluntariness of subsequent confessions. See supra, at 329-332. The Court’s negative implication that Miranda violations are not “improper tactics” is, to say the least, disquieting. Second, the Court reasons that the fact that the accused gave a subsequent confession is itself “highly probative” evidence that he was able to exercise his free will. Ante, at 318. This inaccurate premise follows from the Court’s erroneous rejection of the “cat out of the bag” presumption in these circumstances and its inexplicable assertion that the previous extraction of a “guilty secret” neither constitutes compulsion nor compromises the volun-tariness of later confessions. Ante, at 312.25 Finally, the *344foundation of the derivative-evidence doctrine has always been that, where the authorities have acted illegally, they must bear the “ultimate burden” of proving that their misconduct did not “taint” subsequently obtained evidence. Alderman v. United States, 394 U. S. 165, 183 (1969); see also Nardone v. United States, 308 U. S., at 341. That is precisely the point of the derivative-evidence presumption. By rejecting this presumption in Miranda cases, the Court today appears to adopt a “go ahead and try to prove it” posture toward citizens whose Fifth Amendment Miranda rights have been violated, an attitude that marks a sharp break from the Court’s traditional approach to official lawlessness.

Nevertheless, prudent law enforcement officials must not now believe that they are wholly at liberty to refuse to give timely warnings and obtain effective waivers, confident that evidence derived from Miranda violations will be entirely immune from judicial scrutiny. I believe that most state and federal courts will continue to exercise the “learning, good sense, fairness and courage” they have displayed in administering the derivative-evidence rules prior to today’s decision. Nardone v. United States, supra, at 342. Lower courts are free to interpret the Court’s qualifications, grudging though they may be, as providing sufficient latitude to scrutinize confessions obtained in the wake of Miranda violations to determine whether, in light of all “the surrounding circumstances and the entire course of police conduct,” the initial Miranda violation compromised the voluntariness of the accused’s subsequent confession. Ante, at 318. Any overt *345use of the illegally secured statement by the police in obtaining the subsequent confession must of course be viewed as powerful evidence of a tainted connection; the Court itself asserts that the officers in this case did not “exploit the unwarned admission to pressure respondent” into giving his subsequent confession. Ante, at 316.26 In such circumstances, “[h]aving ‘released the spring’ by using the petitioner’s unlawfully obtained confessions against him, the Government must show that its illegal action did not induce his [subsequent statements].” Harrison v. United States, 392 U. S. 219, 224-225 (1968).

Moreover, courts must scrutinize the totality of the circumstances even where the authorities have not explicitly exploited the earlier confession. Many of the police practices discussed above do not rely on overt use of the earlier confession at all, but instead are implicit strategies that create leverage on the accused to believe he already has sealed his fate. See swpra, at 328-332. These strategies are just as pernicious as overt exploitation of the illegal confession, because they just as surely are “calculated to undermine the suspect’s ability to exercise his free will.” Ante, at 309.27 In evaluating the likely effects of such tactics, courts should continue to employ many of the same, elements traditionally used in dissipation analysis. Thus, although the Court discounts the importance of a “break in the stream of events” in *346the context of the derivative-evidence presumption, the proximity in time and place of the first and second confessions surely remains a critical factor. See supra, at 339-341. So too does the inquiry into possible intervening events. Supra, at 341-342. And if the official violation of Miranda was flagrant, courts may fairly conclude that the violation was calculated and employed precisely so as to “undermine the suspect’s ability to exercise his free will.” Ante, at 309. See also ante, at 314 (“deliberately . . . improper tactics” warrant a presumption of compulsion).28

In sum, today’s opinion marks an evisceration of the established fruit of the poisonous tree doctrine, but its reasoning is sufficiently obscure and qualified as to leave state and federal courts with continued authority to combat obvious flouting by the authorities of the privilege against self-incrimination. I am confident that lower courts will exercise this authority responsibly, as they have for the most part prior to this Court’s intervention.

II

Not content merely to ignore the practical realities of police interrogation and the likely effects of its abolition of the derivative-evidence presumption, the Court goes on to assert that nothing in the Fifth Amendment or the general, judicial policy of deterring illegal police conduct “ordinarily” requires the suppression of evidence derived proximately from a confession obtained in violation of Miranda. The Court does not limit its analysis to successive confessions, but recurrently refers generally to the “fruits” of the illegal confession. Ante, at 306, 307, 308. Thus the potential impact of the Court’s reasoning might extend far beyond the *347“cat out of the bag” context to include the discovery of physical evidence and other derivative fruits of Miranda violations as well.29

A

The Fifth Amendment requires that an accused in custody be informed of important constitutional rights before the authorities interrogate him. Miranda v. Arizona. This requirement serves to combat the “inherently compelling pressures” of custodial questioning “which work to undermine the individual’s will to resist and to compel him to speak where he would not otherwise do so freely,” and is a prerequisite to securing the accused’s informed and voluntary waiver of his *348rights. 384 U. S., at 467. Far from serving merely as a prophylactic safeguard, “[t]he requirement of warnings and waiver of rights is a fundamental with respect to the Fifth Amendment privilege . . . .” Id., at 476. It is precisely because this requirement embraces rights that are deemed to serve a “central role in the preservation of basic liberties,” Malloy v. Hogan, 378 U. S. 1, 5 (1964), that it is binding on the States through the Fourteenth Amendment, Miranda v. Arizona, 384 U. S., at 467.

Twice in the last 10 years, however, the Court has suggested that the Miranda safeguards are not themselves rights guaranteed by the Fifth Amendment. In Michigan v. Tucker, 417 U. S. 433 (1974), the Court stated that Miranda had only prescribed “recommended” procedural safeguards “to provide practical reinforcement for the right against compulsory self-incrimination,” the violation of which may not necessarily violate the Fifth Amendment itself. 417 U. S., at 443-444. And in New York v. Quarles, 467 U. S. 649 (1984), the Court last Term disturbingly rejected "the argument that a confession “must be presumed compelled because of . . . failure to read [the accused] his Miranda warnings.” Id., at 655, n. 5 (emphasis in original).

These assertions are erroneous. Miranda’s requirement of warnings and an effective waiver was not merely an exercise of supervisory authority over interrogation practices. As Justice Douglas noted in his Tucker dissent:

“Miranda’s purpose was not promulgation of judicially preferred standards for police interrogation, a function we are quite powerless to perform; the decision enunciated ‘constitutional standards for protection of the privilege’ against self-incrimination. 384 U. S., at 491.” 417 U. S., at 465-466 (emphasis in original).

Miranda clearly emphasized that warnings and an informed waiver are essential to the Fifth Amendment privilege itself. See supra, at 347 and this page. As noted in Tucker, Miranda did state that the Constitution does not require *349“‘adherence to any particular solution’” for providing the required knowledge and obtaining an informed waiver. 417 U. S., at 444 (quoting Miranda, supra, at 467). But to rely solely on this language in concluding that the Miranda warnings are not constitutional rights, as did the Court in Tucker, ignores the central issue. The Court in Tucker omitted to mention that in Miranda, after concluding that no “particular solution” is required, we went on to emphasize that “unless we are shown other procedures which are at least as effective in apprising accused persons of their right of silence and in assuring a continuous opportunity to exercise it, the [prescribed] safeguards must be observed.” Miranda, supra, at 467. Thus “the use of [any] admissions obtained in the absence of the required warnings [is] a flat violation of the Self-Incrimination Clause of the Fifth Amendment . . . .” Orozco v. Texas, 394 U. S. 324, 326 (1969).

The Court today finally recognizes these flaws in the logic of Tucker and Quarles. 30 Although disastrous in so many other respects, today’s opinion at least has the virtue of rejecting the inaccurate assertion in Quarles that confessions extracted in violation of Miranda are not presumptively coerced for Fifth Amendment purposes. Cf. Quarles, supra, at 655, n. 5. Instead, the Court holds squarely that there is an “irrebuttable” presumption that such confessions are indeed coerced and are therefore inadmissible under the Fifth Amendment except in narrow circumstances. Ante, at 307. 31

Unfortunately, the Court takes away with one hand far more than what it has given with the other. Although the *350Court concedes, as it must, that a confession obtained in violation of Miranda is irrebuttably presumed to be coerced and that the Self-Incrimination Clause therefore prevents its use in the prosecution’s case in chief, ante, at 306-307, the Court goes on to hold that nothing in the Fifth Amendment prevents the introduction at trial of evidence proximately derived from the illegal confession. It contends, for example, that the Fifth Amendment prohibits introduction “only” of the “compelled testimony,” and that this constitutional guarantee “is not concerned with nontestimonial evidence.” Ante, at 304, 307.

This narrow compass of the protection against compelled self-incrimination does not accord with our historic understanding of the Fifth Amendment. Although the Self-Incrimination Clause “protects an accused only from being compelled to testify against himself, or otherwise provide the State with evidence of a testimonial or communicative nature,” Schmerber v. California, 384 U. S. 757, 761 (1966), it prohibits the use of such communications “against” the accused in any way. The Fifth Amendment therefore contains a self-executing rule commanding the exclusion of evidence derived from such communications.32 It bars “the use of compelled testimony, as well as evidence derived directly and indirectly therefrom,” and “prohibits the prosecutorial authorities from using the compelled testimony in any respect.” Kastigar v. United States, 406 U. S. 441, 453 (1972) (emphasis in original). If a coerced statement leads to “sources of information which may supply other means of convicting” the accused, those sources must also be suppressed. Counselman v. Hitchcock, 142 U. S. 547, 586 (1892). Under this constitutional exclusionary rule, the authorities are thus *351“prohibited from making any . . . use of compelled testimony and its fruits” “in connection with a criminal prosecution against” the accused. Murphy v. Waterfront Comm’n, 378 U. S. 52, 79 (1964) (emphasis added).33

In short, the Fifth Amendment’s rule excluding “the use of compelled testimony and evidence derived therefrom is coextensive with the scope of the privilege” against self-incrimination itself. Kastigar v. United States, supra, at 452-453. “The essence of a provision forbidding the acquisition of evidence in a certain way is that not merely evidence so acquired shall not be used before the Court but that it shall not be used at all.” Silverthorne Lumber Co. v. United States, 251 U. S., at 392 (emphasis added). If the authorities were permitted to use an accused’s illegal confession to extract additional confessions or to uncover physical evidence against him, the use of these fruits at trial would violate the Self-Incrimination Clause just as surely as if the original confession itself were introduced. Yet that is precisely what today’s decision threatens to encourage.

What possible justification does the Court advance for its evisceration of the Fifth Amendment’s exclusionary rule in this context? Two rationales appear to be at work here. First, while acknowledging that a confession obtained in the absence of warnings and an informed waiver is irrebuttably presumed to be coerced in violation of the Self-Incrimination Clause, ante, at 307, the Court recurrently asserts elsewhere that the extraction of such a confession is not really “a Fifth Amendment violation,” ante, at 306. Thus the Court suggests that a Miranda violation does not constitute “police *352infringement of a constitutional right,” that it is not “a constitutional violation,” that a suspect in such circumstances “suffer[s] no identifiable constitutional harm,” and that his “Fifth Amendment rights” have not “actually [been] violated.” Ante, at 304, 305, 307, 316. Similarly, the Court persists in reasoning that a confession obtained in violation of Miranda “ordinarily” should be viewed as “voluntary,” a “voluntary disclosure of a guilty secret,” “freely given,” “non-coerc[ed],” and “wholly voluntary.” Ante, at 311, 312, 318. I have already demonstrated the fallacy of this reasoning. See Part II-A, supra. Suffice it to say that the public will have understandable difficulty in comprehending how a confession obtained in violation of Miranda can at once be (1) “irrebuttabl[y]” presumed to be the product of official compulsion, and therefore suppressible as a matter of federal constitutional law, ante, at 307, 317, and (2) “noncoerc[ed]” and “wholly voluntary,” ante, at 312, 318.

Second, while not discussed in today’s opinion, Justice O’Connor has recently argued that the Fifth Amendment’s exclusion of derivative evidence extends only to confessions obtained when the accused is compelled “to appear before a court, grand jury, or other such formal tribunal,” and not merely when he is “subject to informal custodial police interrogation.” New York v. Quarles, 467 U. S., at 670 (O’Connor, J., concurring in part in judgment and dissenting in part). An accused in this situation, it is argued, “has a much less sympathetic case for obtaining the benefit of a broad suppression ruling.” Ibid.

Such an analysis overlooks that, by the time we decided Miranda, it was settled that the privilege against self-incrimination applies with full force outside the chambers of “formal” proceedings. “Today, then, there can be no doubt that the Fifth Amendment privilege is available outside of criminal court proceedings and serves to protect persons in all settings in which their freedom of action is curtailed in any significant way from being compelled to incriminate themselves.” Miranda v. Arizona, 384 U. S., at 467. See also *353 Ziang Sung Wan v. United States, 266 U. S. 1, 14-15 (1924) (“[A] confession obtained by compulsion must be excluded whatever may have been the character of the compulsion, and whether the compulsion was applied in a judicial proceeding or otherwise”) (emphasis added); Bram v. United States, 168 U. S. 532 (1897). Thus there is no question that “all the principles embodied in the privilege apply to informal compulsion exerted by law-enforcement officers during in-custody questioning.” Miranda v. Arizona, supra, at 461.

The application of the privilege to custodial interrogation simply reflects the realities and purposes of 20th-century police investigations, matters which the Court chooses to ignore. “[PJolice interrogation has in recent times performed the function once accomplished by interrogation of the defendant by the committing magistrate, a practice brought to an end by establishment of the rule against self-incrimination.”34 Moreover, “[a]s a practical matter, the compulsion to speak in the [police interrogation setting] may well be greater than in courts or other official investigations, where there are often impartial observers to guard against intimidation or trickery.” 384 U. S., at 461 (emphasis added).35 In addition, there can be no legitimate dispute that *354an incriminating statement obtained through custodial interrogation “is as revealing of leads” and other derivative evidence as a statement compelled before a judicial tribunal. Murphy v. Waterfront Comm’n, 378 U. S., at 103 (White, J., concurring). Accordingly, Miranda itself emphasized that, under the Fifth Amendment exclusionary rule, “no evidence obtained as a result of interrogation can be used against” the defendant unless he was warned of his rights and gave an effective waiver. 384 U. S., at 479 (emphasis added).36

For these reasons, the Fifth Amendment itself requires the exclusion of evidence proximately derived from a confession obtained in violation of Miranda. The Court today has altogether evaded this constitutional command, the application of which should not turn simply on whether one is “sympathetic” to suspects undergoing custodial interrogation.

C

Even if I accepted the Court’s conclusion that the Fifth Amendment does not command the suppression of evidence proximately derived from a Miranda violation, I would nevertheless dissent from the Court’s refusal to recognize the importance of deterring Miranda violations in appropriate circumstances. Just last Term, in United States v. Leon, 468 U. S. 897 (1984), the Court held that while the Fourth Amendment does not per se require the suppression of evidence derived from an unconstitutional search, the exclusionary rule must nevertheless be invoked where the search was objectively unreasonable. Id., at 919-920, n. 20. Although *355I do not share the Court’s view of the Fourth Amendment,37 Leon at least had the virtue of recognizing that exclusion of derivative evidence is essential to the effective deterrence of objectively unreasonable failures by the authorities to obey the law. Ibid.

The Court today refuses to apply the derivative-evidence rule even to the extent necessary to deter objectively unreasonable failures by the authorities to honor a suspect’s Miranda rights. Incredibly, faced with an obvious violation of Miranda, the Court asserts that it will not countenance suppression of a subsequent confession in such circumstances where the authorities have acted “legitimately]” and have not used “improper tactics.” Ante, at 312, 314. One can only respond: whither went Miranda?

The Court contends, however, that Michigan v. Tucker, 417 U. S. 433 (1974), already decided that the failure of the authorities to obey Miranda should not be deterred by application of the derivative-evidence rule. Ante, at 308-309. Tucker did not so decide. After criticizing the Fifth Amendment basis for exclusion, the Court in Tucker went on to note another “ ‘prime purpose’ ” for the exclusion of evidence — “ ‘to deter future unlawful police conduct and thereby effectuate the guarantee^]’” of the Constitution. 417 U. S., at 446 (citation omitted). The Court emphasized that “[i]n a proper case this rationale would seem applicable to the Fifth Amendment context as well.” Id., at 447. Anticipating Leon, however, the Court asserted that the “deterrent purpose” was applicable only where “the police have engaged in willful, or at the very least negligent, conduct . . . .” 417 U. S, at 447. Because the questioning in Tucker occurred before Miranda was announced and was otherwise conducted in an objectively reasonable manner, the exclusion of the derivative evidence solely for failure to comply with the then-*356nonexistent Miranda requirement would not significantly deter future Miranda violations. As the Court noted, the “deterrence rationale loses much of its force” when there is nothing to deter. 417 U. S, at 447.

Far from rejecting the derivative-evidence rule, Tucker thus expressly invited its application in “a proper case” when the authorities have acted unreasonably. Ibid. Nearly every court and commentator considering the issue have correctly recognized that Tuckers logic and its reliance on the Fourth Amendment “good faith” analysis compel the exclusion of derivative evidence where the police have deliberately, recklessly, or negligently violated the Fifth Amendment requirement of warnings and an effective waiver.38

Thus the Court’s assertion today that Tucker's “reasoning applies with equal force” to preclude application of the derivative-evidence rule in this case is a gross mis-characterization. Ante, at 308. If the police acted in an objectively unreasonable manner, see Part II-D, infra, Tucker’s “reasoning” instead requires suppression of Elstad’s subsequent statement.

The Court clearly errs in suggesting that suppression of the “unwarned admission” alone will provide meaningful deterrence. Ante, at 309. The experience of lower courts demonstrates that the police frequently have refused to comply with Miranda precisely in order to obtain incriminating statements that will undermine the voluntariness of the accused’s decision to speak again once he has received the usual warnings; in such circumstances, subsequent confes*357sions often follow on a “silver platter.” Cagle v. State, 45 Ala. App., at 4, 221 So. 2d, at 120. See generally supra, at 329-332. Expert interrogators themselves recognize the direct connection between such statements. Supra, at 328-329. And the Court’s suggestion that its analysis might apply generally to “fruits” of illegal interrogations, but see n. 29, supra, blinks reality even further. For example, expert interrogators acknowledge that confessions are “‘the prime source of other evidence.’ ”39 If the police through illegal interrogation could discover contraband and be confident that the contraband “ordinarily” would not be suppressed, what possible incentive would they have to obey Miranda?

The Court simply has not confronted the basic premise of the derivative-evidence rule: that “[t]o forbid the direct use of methods . . . but to put no curb on their full indirect use would only invite the very methods deemed ‘inconsistent with ethical standards and destructive of personal liberty.’” Nardone v. United States, 308 U. S., at 340.

“[I]t is clear that if the police were permitted to utilize illegally obtained confessions for links and leads rather than being required to gather evidence independently, then the Miranda warnings would be of no value in protecting the privilege against self-incrimination. The requirement of a warning would be meaningless, for the police would be permitted to accomplish indirectly what they could not accomplish directly, and there would exist no incentive to warn.” Pitler, 56 Calif. L. Rev., supra n. 16, at 620.

*358As the Executive Director of the National District Attorneys Association Foundation emphasized shortly after Miranda, merely to exclude the statement itself while putting no curbs on the admission of derivative evidence “would destroy the whole basis for the rule in the first instance.” Nedrud, The New Fifth Amendment Concept: Self-Incrimination Redefined, 2 J. Nat. Dist. Att. Assn. Found. 112,114 (1966).40 Yet that is precisely the result that today’s disastrous opinion threatens to encourage. How can the Court possibly expect the authorities to obey Miranda when they have every incentive now to interrogate suspects without warnings or an effective waiver, knowing that the fruits of such interrogations “ordinarily” will be admitted, that an admissible subsequent confession “ordinarily” can be obtained simply by reciting the Miranda warnings shortly after the first has been procured and asking the accused to repeat himself, and that unless the accused can demonstrate otherwise his confession will be viewed as an “act of free will” in response to “legitimate law *359enforcement activity”? Ante, at 311, 312. By condoning such a result, the Court today encourages practices that threaten to reduce Miranda to a mere “form of words,” Silverthorne Lumber Co. v. United States, 251 U. S., at 392, and it is shocking that the Court nevertheless disingenuously purports that it “in no way retreats” from the Miranda safeguards, ante, at 317.

D

Not content with its handiwork discussed above, the Court goes on and devotes considerable effort to suggesting that, “[u]nfortunately,” Miranda is such an inherently “slippery,” “murky,” and “difficult” concept that the authorities in general, and the police officer conducting the interrogation in this case in particular, cannot be faulted for failing to advise a suspect of his rights and to obtain an informed waiver. Ante, at 309, 316. Miranda will become “murky,” however, only because the Court’s opinion today threatens to become a self-fulfilling prophecy. Although borderline cases occasionally have arisen respecting the concepts of “custody” and “interrogation,” until today there has been nothing “slippery,” “murky,” or “difficult” about Miranda in the overwhelming majority of cases. The whole point of the Court’s work in this area has been to prescribe “bright line” rules to give clear guidance to the authorities.41

Rather than acknowledge that the police in this case clearly broke the law, the Court bends over backwards to suggest why the officers may have been justified in failing to obey Miranda.

*360 First. The Court asserts that “[njeither the environment nor the manner of either ‘interrogation’ was coercive,” noting that the initial interrogation took place in Elstad’s “own home.” Ante, at 315. The Court also believes that, “[a]l-though in retrospect the officers testified that respondent was then in custody, at the time he made his statement he had not been informed that he was under arrest.” Ibid. There is no question, however, that Michael Elstad was in custody and “deprived of his freedom of action in [a] significant way” at the time he was interrogated. Miranda v. Arizona, 384 U. S., at 444. Two police officers had entered his bedroom, ordered him to get out of bed and come with them, stood over him while he dressed, taken him downstairs, and separated him from his mother. Tr. 64-65, 74-75, 80-84. The officers themselves acknowledged that Elstad was then under arrest. Id., at 81-82. Moreover, we have made clear that police interrogation of an accused in custody triggers the Miranda safeguard even if he is in the “familiar surroundings” of his own home, precisely because he is no less “ ‘deprived of his freedom of action’ ” there than if he were at a police station. Orozco v. Texas, 394 U. S., at 326-327 (citation omitted).

Thus because Elstad was in custody, the circumstances of his interrogation were inherently coercive, and the Court once again flouts settled law in suggesting otherwise. “[WJithout proper safeguards the process of in-custody interrogation of persons suspected or accused of crime contains inherently compelling pressures which work to undermine the individual’s will to resist and to compel him to speak where he would not otherwise do so freely.” Miranda v. Arizona, 384 U. S., at 467. The Fifth Amendment’s requirement of warnings and an informed waiver is “an absolute prerequisite in overcoming the inherent pressures of the interrogation atmosphere.” Id., at 468.

Second. Without anything in the record to support its speculation, the Court suggests that Officer Burke’s violation *361of Miranda “may have been the result of confusion as to whether the brief exchange qualified as ‘custodial interrogation’ . . . Ante, at 315. There was no confusion on this point until today. Burke made Elstad sit down and, standing over him, said “[y]ou know why we’re here,” asked if he knew the Gross family, and “asked what he knew about the burglary.” Tr. 83-84. This questioning obviously constituted interrogation because it was “reasonably likely to evoke an incriminating response” from Elstad, as it did. Rhode Island v. Innis, 446 U. S. 291, 301 (1980).

Third. The Court contends that the interrogation might be excusable because “the brief stop in the living room before proceeding to the station house was not to interrogate the suspect but to notify his mother of the reason for his arrest.” Ante, at 315. Officer Burke’s partner did take Elstad’s mother into the kitchen to inform her of the charges, but Burke took Elstad into another room, sat him down, and interrogated him concerning “what he knew about the burglary.” Tr. 84. How can the Court possibly describe this interrogation as merely informing Elstad’s mother of his arrest?

Finally. The Court suggests that Burke’s violation of Elstad’s Fifth Amendment rights “may simply have reflected Burke’s reluctance to initiate an alarming police procedure before McAllister had spoken with respondent’s mother.” Ante, at 315-316. As the officers themselves acknowledged, however, the fact that they “[took] the young fellow out of bed” had “[ojbviously” already created “tension and stress” for the mother, Tr. 64, which surely was not lessened when she learned that her son was under arrest. And if Elstad’s mother was in earshot, as the Court assumes, it is difficult to perceive how listening to the Miranda warnings would be any more “alarming” to her than what she actually heard— actual interrogation of her son, including Burke’s direct accusation that the boy had committed a felony. Most importantly, an individual’s constitutional rights should not turn on *362whether his relatives might be upset. Surely there is no “tender feelings” exception to the Fifth Amendment privilege against self-incrimination.42

I — I hH I — I

The Court’s decision today vividly reflects its impatience with the constitutional rights that the authorities attack as standing in the way of combating crime. But the States that adopted the Bill of Rights struck that balance and it is not for this Court to balance the Bill of Rights away on a cost/benefit scale “where the ‘costs’ of excluding illegally obtained evidence loom to exaggerated heights and where the ‘benefits’ of such exclusion are made to disappear with a mere wave of the hand.” United States v. Leon, 468 U. S., at 929 (Brennan, J., dissenting). It is precisely in that vein, however, that the Court emphasizes that the subsequent confession in this case was “voluntary” and “highly probative evidence,” that application of the derivative-evidence presumption would cause the confession to be “irretrievably lost,” and that such a result would come at an impermissibly “high cost to legitimate law enforcement activity.” Ante, at 312.

Failure of government to obey the law cannot ever constitute “legitimate law enforcement activity.” In any event, application of the derivative-evidence presumption does not *363“irretrievably” lead to suppression. If a subsequent confession is truly independent of earlier, illegally obtained confessions, nothing prevents its full use to secure the accused’s conviction. If the subsequent confession did result from the earlier illegalities, however, there is nothing “voluntary” about it. And even if a tainted subsequent confession is “highly probative,” we have never until today permitted probity to override the fact that the confession was “the product of constitutionally impermissible methods in [its] inducement.” Rogers v. Richmond, 365 U. S. 534, 541 (1961). In such circumstances, the Fifth Amendment makes clear that the prosecutor has no entitlement to use the confession in attempting to obtain the accused’s conviction.43

The lesson of today’s decision is that, at least for now, what the Court decrees are “legitimate” violations by authorities of the rights embodied in Miranda shall “ordinarily” go undeterred. It is but the latest of the escalating number of decisions that are making this tribunal increasingly irrelevant in the protection of individual rights, and that are requiring other tribunals to shoulder the burden.44 “There is hope, however, that in time this or some later Court will restore *364these precious freedoms to their rightful place as a primary protection for our citizens against overreaching officialdom.” United States v. Leon, supra, at 960 (Brennan, J., dissenting).

I dissent.

Justice Stevens,

dissenting.

The Court concludes its opinion with a carefully phrased statement of its holding:

“We hold today that a suspect who has once responded to unwarned yet uncoercive questioning is not thereby disabled from waiving his rights and confessing after he has been given the requisite Miranda warnings.” Ante, at 318.

I find nothing objectionable in such a holding. Moreover, because the Court expressly endorses the “bright-line rule of Miranda,” which conclusively presumes that incriminating statements obtained from a suspect in custody without administering the required warnings are the product of compulsion,1 and because the Court places so much emphasis on the special facts of this case, I am persuaded that the Court intends its holding to apply only to a narrow category of cases in which the initial questioning of the suspect was made in a totally uncoercive setting and in which the first confession obviously had no influence on the second.2 I nevertheless *365dissent because even such a narrowly confined exception is inconsistent with the Court’s prior cases, because the attempt to identify its boundaries in future cases will breed confusion and uncertainty in the administration of criminal justice, and because it denigrates the importance of one of the core constitutional rights that protects every American citizen from the kind of tyranny that has flourished in other societies.

I

The desire to achieve a just result in this particular case has produced an opinion that is somewhat opaque and internally inconsistent. If I read it correctly, its conclusion rests on two untenable premises: (1) that the respondent’s first confession was not the product of coercion;3 and (2) that no constitutional right was violated when respondent was questioned in a tranquil, domestic setting.4

*366Even before the decision in Miranda v. Arizona, 384 U. S. 436 (1966), it had been recognized that police interrogation of a suspect who has been taken into custody is presumptively coercive. That presumption had its greatest force when the questioning occurred in a police station, when it was prolonged, and when there was evidence that the prisoner had suffered physical injury. To rebut the presumption, the prosecutor had the burden of proving the absence of any actual coercion.5 Because police officers are generally more credible witnesses than prisoners and because it is always difficult for triers of fact to disregard evidence of guilt when addressing a procedural question, more often than not the presumption of coercion afforded only slight protection to the accused.

The decision in Miranda v. Arizona clarified the law in three important respects. First, it provided the prosecutor with a simple method of overcoming the presumption of coercion.6 If the police interrogation is preceded by the warning specified in that opinion, the usual presumption does not attach. Second, it provided an important protection to the accused by making the presumption of coercion irrebuttable if the prescribed warnings are not given.7 Third, the decision *367made it clear that a self-incriminatory statement made in response to custodial interrogation was always to be considered “compelled” within the meaning of the Fifth Amendment to the Federal Constitution if the interrogation had not been preceded by appropriate warnings.8 Thus the irrebuttable presumption of coercion that applies to such a self-incriminatory statement, like a finding of actual coercion, renders the resulting confession inadmissible as a matter of federal constitutional law.9

*368In my opinion, the Court’s attempt to fashion a distinction between actual coercion “by physical violence or other deliberate means calculated to break the suspect’s will,” ante, at 312, and irrebuttably presumed coercion cannot succeed. The presumption is only legitimate if it is assumed that there is always a coercive aspect to custodial interrogation that is not preceded by adequate advice of the constitutional right to remain silent. Although I would not support it, I could understand a rule that refused to apply the presumption unless the interrogation took place in an especially coercive setting — perhaps only in the police station itself — but if the presumption arises whenever the accused has been taken into custody or his freedom has been restrained in any significant way, it will surely be futile to try to develop subcategories of custodial interrogation.10 Indeed, a major purpose of treat-' ing the presumption of coercion as irrebuttable is to avoid the kind of fact-bound inquiry that today’s decision will surely engender.11

As I read the Court’s opinion, it expressly accepts the proposition that routine Miranda warnings will not be sufficient to overcome the presumption of coercion and thereby make a second confession admissible when an earlier confession is tainted by coercion “by physical violence or other *369deliberate means calculated to break the suspect’s will.”12 Even in such a case, however, it is not necessary to assume that the earlier confession will always “effectively immunize” a later voluntary confession. But surely the fact that an earlier confession was obtained by unlawful methods should add force to the presumption of coercion that attaches to subsequent custodial interrogation and should require the prosecutor to shoulder a heavier burder of rebuttal than in a routine case. Simple logic, as well as the interest in not providing an affirmative incentive to police misconduct, requires that result. I see no reason why the violation of a rule that is as well recognized and easily administered as the duty to give Miranda warnings should not also impose an additional burden on the prosecutor.13 If we are faithful to the holding in *370 Miranda itself, when we are considering the admissibility of evidence in the prosecutor’s case in chief, we should not try to fashion a distinction between police misconduct that warrants a finding of actual coercion and police misconduct that establishes an irrebuttable presumption of coercion.

H-I HH

For me, the most disturbing aspect of the Court’s opinion is its somewhat opaque characterization of the police misconduct in this case. The Court appears ambivalent on the question whether there was any constitutional violation.14 This ambivalence is either disingenuous or completely lawless. This Court’s power to require state courts to exclude probative self-incriminatory statements rests entirely on the premise that the use of such evidence violates the Federal Constitution.15 The same constitutional analysis applies *371whether the custodial interrogation is actually coercive or irrebuttably presumed to be coercive. If the Court does not accept that premise, it must regard the holding in the Miranda case itself, as well as all of the federal jurisprudence that has evolved from that decision, as nothing more than an illegitimate exercise of raw judicial power.16 , If the Court accepts the proposition that respondent’s self-incriminatory statement was inadmissible, it must also acknowledge that the Federal Constitution protected him from custodial police interrogation without first being advised of his right to remain silent.

The source of respondent’s constitutional protection is the Fifth Amendment’s privilege against compelled self-incrimination that is secured against state invasion by the Due Process Clause of the Fourteenth Amendment. Like many other provisions of the Bill of Rights, that provision is merely a procedural safeguard. It is, however, the specific provision that protects all citizens from the kind of custodial interrogation that was once employed by the Star Chamber,17 by “the Germans of the 1930’s and early 1940’s,” 18 and by some of our own police departments only a few decades ago.19 *372Custodial interrogation that violates that provision of the Bill of Rights is a classic example of a violation of a constitutional right.

I respectfully dissent.

*604Justice Souter

announced the judgment of the Court and delivered an opinion, in which Justice Stevens, Justice Ginsburg, and Justice Breyer join.

This case tests a police protocol for custodial interrogation that calls for giving no warnings of the rights to silence and counsel until interrogation has produced a confession. Although such a statement is generally inadmissible, since taken in violation of Miranda v. Arizona, 384 U. S. 436 (1966), the interrogating officer follows it with Miranda warnings and then leads the suspect to cover the same ground a second time. The question here is the admissibility of the repeated statement. Because this midstream recitation of warnings after interrogation and unwarned confession could not effectively comply with Miranda’s constitutional requirement, we hold that a statement repeated after a warning in such circumstances is inadmissible.

I

Respondent Patrice Seibert’s 12-year-old son Jonathan had cerebral palsy, and when he died in his sleep she feared charges of neglect because of bedsores on his body. In her presence, two of her teenage sons and two of their friends devised a plan to conceal the facts surrounding Jonathan’s death by incinerating his body in the course of burning the family’s mobile home, in which they planned to leave Donald Rector, a mentally ill teenager living with the family, to avoid any appearance that Jonathan had been unattended. Seibert’s son Darían and a friend set the fire, and Donald died.

Five days later, the police awakened Seibert at 3 a.m. at a hospital where Darían was being treated for burns. In arresting her, Officer Kevin Clinton followed instructions from Rolla, Missouri, Officer Richard Hanrahan that he refrain from giving Miranda warnings. After Seibert had been taken to the police station and left alone in an interview room for 15 to 20 minutes, Officer Hanrahan questioned her *605without Miranda warnings for 30 to 40 minutes, squeezing her arm and repeating “Donald was also to die in his sleep.” App. 59 (internal quotation marks omitted). After Seibert finally admitted she knew Donald was meant to die in the fire, she was given a 20-minute coffee and cigarette break. Officer Hanrahan then turned on a tape recorder, gave Seibert the Miranda warnings, and obtained a signed waiver of rights from her. He resumed the questioning with “Ok, ’trice, we’ve been talking for a little while about what happened on Wednesday the twelfth, haven’t we?” App. 66, and confronted her with her prewarning statements:

Hanrahan: “Now, in discussion you told us, you told us that there was a[n] understanding about Donald.”

Seibert: “Yes.”

Hanrahan: “Did that take place earlier that morning?” Seibert: “Yes.”

Hanrahan: “And what was the understanding about Donald?”

Seibert: “If they could get him out of the trailer, to take him out of the trailer.”

Hanrahan: “And if they couldn’t?”

Seibert: “I, I never even thought about it. I just figured they would.”

Hanrahan: “’Trice, didn’t you tell me that he was supposed to die in his sleep?”

Seibert: “If that would happen, ’cause he was on that new medicine, you know ....”

Hanrahan: “The Prozac? And it makes him sleepy. So he was supposed to die in his sleep?”

Seibert: “Yes.” Id., at 70.

After being charged with first-degree murder for her role in Donald’s death, Seibert sought to exclude both her pre-warning and postwarning statements. At the suppression hearing, Officer Hanrahan testified that he made a “conscious *606decision” to withhold Miranda warnings, thus resorting to an interrogation technique he had been taught: question first, then give the warnings, and then repeat the question “until I get the answer that she’s already provided once.” App. 31-34. He acknowledged that Seibert’s ultimate statement was “largely a repeat of information . . . obtained” prior to the warning. Id., at 30.

The trial court suppressed the prewarning statement but admitted the responses given after the Miranda recitation. A jury convicted Seibert of second-degree murder. On appeal, the Missouri Court of Appeals affirmed, treating this case as indistinguishable from Oregon v. Elstad, 470 U. S. 298 (1985). No. 23729, 2002 WL 114804 (Jan. 30, 2002) (not released for publication).

The Supreme Court of Missouri reversed, holding that “[i]n the circumstances here, where the interrogation was nearly continuous, . . . the second statement, clearly the product of the invalid first statement, should have been suppressed.” 93 S. W. 3d 700, 701 (2002) (en banc). The court distinguished Elstad on the ground that warnings had not intentionally been withheld there, 93 S. W. 3d, at 704, and reasoned that “Officer Hanrahan’s intentional omission of a Miranda warning was intended to deprive Seibert of the opportunity knowingly and intelligently to waive her Miranda rights,” id., at 706. Since there were “no circumstances that would seem to dispel the effect of the Miranda violation,” the court held that the postwarning confession was involuntary and therefore inadmissible. Ibid. To allow the police to achieve an “end run” around Miranda, the court explained, would encourage Miranda violations and diminish Miranda’s role in protecting the privilege against self-incrimination. 93 S. W. 3d, at 706-707. Three judges dissented, taking the view that Elstad applied even though the police intentionally withheld Miranda warnings before the initial statement, and believing that “Seibert’s unwarned responses to Officer Hanrahan’s questioning did not prevent *607her from waiving her rights and confessing.” 93 S. W. 3d, at 708 (opinion of Benton, J.).

We granted certiorari, 538 U. S. 1031 (2003), to resolve a split in the Courts of Appeals. Compare United States v. Gale, 952 F. 2d 1412, 1418 (CADC 1992) (while “deliberate ‘end run’ around Miranda” would provide cause for suppression, case involved no conduct of that order); United States v. Carter, 884 F. 2d 368, 373 (CA8 1989) (“Elstad did not go so far as to fashion a rule permitting this sort of end run around Miranda”), with United States v. Orso, 266 F 3d 1030, 1034-1039 (CA9 2001) (en banc) (rejecting argument that “tainted fruit” analysis applies because deliberate withholding of Miranda warnings constitutes an “improper tactic”); United States v. Esquilin, 208 F 3d 315, 319-321 (CA1 2000) (similar). We now affirm.

II

“In criminal trials, in the courts of the United States, wherever a question arises whether a confession is incompetent because not voluntary, the issue is controlled by that portion of the Fifth Amendment . . . commanding that no person ‘shall be compelled in any criminal case to be a witness against himself.’ ” Bram v. United States, 168 U. S. 532, 542 (1897). A parallel rule governing the admissibility of confessions in state courts emerged from the Due Process Clause of the Fourteenth Amendment, see, e. g., Brown v. Mississippi, 297 U. S. 278 (1936), which governed state cases until we concluded in Malloy v. Hogan, 378 U. S. 1, 8 (1964), that “[t]he Fourteenth Amendment secures against state invasion the same privilege that the Fifth Amendment guarantees against federal infringement — the right of a person to remain silent unless he chooses to speak in the unfettered exercise of his own will, and to suffer no penalty ... for such silence.” In unifying the Fifth and Fourteenth Amendment voluntariness tests, Malloy “made clear what had already become apparent — that the substantive and procedural safe*608guards surrounding admissibility of confessions in state cases had become exceedingly exacting, reflecting all the policies embedded in the privilege” against self-incrimination. Miranda, 384 U. S., at 464.

In Miranda, we explained that the “voluntariness doctrine in the state cases ... encompasses all interrogation practices which are likely to exert such pressure upon an individual as to disable him from making a free and rational choice,” id., at 464-465. We appreciated the difficulty of judicial enquiry post hoc into the circumstances of a police interrogation, Dickerson v. United States, 530 U. S. 428, 444 (2000), and recognized that “the coercion inherent in custodial interrogation blurs the line between voluntary and involuntary statements, and thus heightens the risk” that the privilege against self-incrimination will not be observed, id., at 435. Hence our concern that the “traditional totality-of-the-eircumstances” test posed an “unacceptably great” risk that involuntary custodial confessions would escape detection. Id., at 442.

Accordingly, “to reduce the risk of a coerced confession and to implement the Self-Incrimination Clause,” Chavez v. Martinez, 538 U. S. 760, 790 (2003) (Kennedy, J., concurring in part and dissenting in part), this Court in Miranda concluded that “the accused must be adequately and effectively apprised of his rights and the exercise of those rights must be fully honored,” 384 U. S., at 467. Miranda conditioned the admissibility at trial of any custodial confession on warning a suspect of his rights: failure to give the prescribed warnings and obtain a waiver of rights before custodial questioning generally requires exclusion of any statements obtained.1 Conversely, giving the warnings and getting a *609waiver has generally produced a virtual ticket of admissibility; maintaining that a statement is involuntary even though given after warnings and voluntary waiver of rights requires unusual stamina, and litigation over voluntariness tends to end with the finding of a valid waiver. See Berkemer v. McCarty, 468 U. S. 420, 433, n. 20 (1984) (“[C]ases in which a defendant can make a colorable argument that a self-incriminating statement was ‘compelled’ despite the fact that the law enforcement authorities adhered to the dictates of Miranda are rare”). To point out the obvious, this common consequence would not be common at all were it not that Miranda warnings are customarily given under circumstances allowing for a real choice between talking and remaining silent.

Ill

There are those, of course, who preferred the old way of doing things, giving no warnings and litigating the voluntariness of any statement in nearly every instance. In the aftermath of Miranda, Congress even passed a statute seeking to restore that old regime, 18 U. S. C. §3501, although the Act lay dormant for years until finally invoked and challenged in Dickerson v. United States, swpra. Dickerson reaffirmed Miranda and held that its constitutional character prevailed against the statute.

The technique of interrogating in successive, unwarned' and warned phases raises a new challenge to Miranda. Although we have no statistics on the frequency of this practice, it is not confined to Rolla, Missouri. An officer of that police department testified that the strategy of withholding Miranda warnings until after interrogating and drawing out a confession was promoted not only by his own department, but by a national police training organization and other departments in which he had worked. App. 31-32. Consistently with the officer’s testimony, the Police Law Institute, for example, instructs that “officers may conduct a two-stage interrogation. ... At any point during the pre-Miranda in*610terrogation, usually after arrestees have confessed, officers may then read the Miranda warnings and ask for a waiver. If the arrestees waive their Miranda rights, officers will be able to repeat any subsequent incriminating statements later in court.” Police Law Institute, Illinois Police Law Manual 83 (Jan. 2001-Dec. 2003) (available in Clerk of Court's case file) (hereinafter Police Law Manual) (emphasis in original).2 *611The upshot of all this advice is a question-first practice of some popularity, as one can see from the reported cases describing its use, sometimes in obedience to departmental policy.3

IV

When a confession so obtained is offered and challenged, attention must be paid to the conflicting objects of Miranda and question-first. Miranda addressed “interrogation practices . . . likely ... to disable [an individual] from making a free and rational choice” about speaking, 384 U. S., at 464-465, and held that a suspect .must be “adequately and effectively” advised of the choice the Constitution guarantees, id., at 467. The object of question-first is to render Miranda warnings ineffective by waiting for a particularly opportune time to give them, after the suspect has already confessed.

Just as “no talismanic incantation [is] required to satisfy [Miranda’s] strictures,” California v. Prysock, 453 U. S. 355, 359 (1981) (per curiam), it would be absurd to think that mere recitation of the litany suffices to satisfy Miranda in every conceivable circumstance. “The inquiry is simply whether the warnings reasonably ‘conve[y] to [a suspect] his rights as required by Miranda.’ ” Duckworth v. Eagan, 492 U. S. 195, 203 (1989) (quoting Prysock, supra, at 361). The threshold issue when interrogators question first and warn later is thus whether it would be reasonable to find that in these circumstances the warnings could function “effec*612tively” as Miranda requires. Could the warnings effectively advise the suspect that he had a real choice about giving an admissible statement at that juncture? Could they reasonably convey that he could choose to stop talking even if he had talked earlier? For unless the warnings could place a suspect who has just been interrogated in a position to make such an informed choice, there is no practical justification for accepting the formal warnings as compliance with Miranda, or for treating the second stage of interrogation as distinct from the first, unwarned and inadmissible segment.4

There is no doubt about the answer that proponents of question-first give to this question about the effectiveness of *613warnings given only after successful interrogation, and we think their answer is correct. By any objective measure, applied to circumstances exemplified here, it is likely that if the interrogators employ the technique of withholding warnings until after interrogation succeeds in eliciting a confession, the warnings will be ineffective in preparing the suspect for successive interrogation, close in time and similar in content. After all, the reason that question-first is catching on is as obvious as its manifest purpose, which is to get a confession the suspect would not make if he understood his rights at the outset; the sensible underlying assumption is that with one confession in hand before the warnings, the interrogator can count on getting its duplicate, with trifling additional trouble. Upon hearing warnings only in the aftermath of interrogation and just after making a confession, a suspect would hardly think he had a genuine right to remain silent, let alone persist in so believing once the police began to lead him over the same ground again.5 A more likely reaction on a suspect’s part would be perplexity about the reason for discussing rights at that point, bewilderment being an unpromising frame of mind for knowledgeable decision. What is worse, telling a suspect that “anything you say can and will be used against you,” without expressly excepting the statement just given, could lead to an entirely reasonable inference that what he has just said will be used, with subsequent silence being of no avail. Thus, when Miranda warnings are inserted in the midst of coordinated and continuing interrogation, they are likely to mislead and “de*614priv[e] a defendant of knowledge essential to his ability to understand the nature of his rights and the consequences of abandoning them.” Moran v. Burbine, 475 U. S. 412, 424 (1986). By the same token, it would ordinarily be unrealistic to treat two spates of integrated and proximately conducted questioning as independent interrogations subject to independent evaluation simply because Miranda warnings formally punctuate them in the middle.

V

Missouri argues that a confession repeated at the end of an interrogation sequence envisioned in a question-first strategy is admissible on the authority of Oregon v. Elstad, 470 U. S. 298 (1985), but the argument disfigures that case. In Elstad, the police went to the young suspect’s house to take him into custody on a charge of burglary. Before the arrest, one officer spoke with the suspect’s mother, while the other one joined the suspect in a “brief stop in the living room,” id., at 315, where the officer said he “felt” the young man was involved in a burglary, id., at 301 (internal quotation marks omitted). The suspect acknowledged he had been at the scene. Ibid. This Court noted that the pause in the living room “was not to interrogate the suspect but to notify his mother of the reason for his arrest,” id., at 315, and described the incident as having “none of the earmarks of coercion,” id., at 316. The Court, indeed, took care to mention that the officer's initial failure to warn was an “oversight” that “may have been the result of confusion as to whether the brief exchange qualified as 'custodial interrogation’ or . . . may simply have reflected . . . reluctance to initiate an alarming police procedure before [an officer] had spoken with respondent’s mother." Id., at 315-316. At the outset of a later and systematic station house interrogation going well beyond the scope of the laconic prior admission, the suspect was given Miranda warnings and made a full confession. Elstad, supra, at 301, 314-315. In holding the *615second statement admissible and voluntary, Elstad rejected the “cat out of the bag” theory that any short, earlier admission, obtained in arguably innocent neglect of Miranda, determined the character of the later, warned confession, El-stad, 470 U. S., at 311-314; on the facts of that case, the Court thought any causal connection between the first and second responses to the police was “speculative and attenuated,” id., at 313. Although the Elstad Court expressed no explicit conclusion about either officer’s state of mind, it is fair to read Elstad as treating the living room conversation as a good-faith Miranda mistake, not only open to correction by careful warnings before systematic questioning in that particular case, but posing no threat to warn-first practice generally. See Elstad, supra, at 309 (characterizing the officers’ omission of Miranda warnings as “a simple failure to administer the warnings, unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will”); 470 U. S., at 318, n. 5 (Justice Brennan’s concern in dissent that Elstad would invite question-first practice “distorts the reasoning and holding of our decision, but, worse, invites trial courts and prosecutors to do the same”).

The contrast between Elstad and this case reveals a series of relevant facts that bear on whether Miranda warnings delivered midstream could be effective enough to accomplish their object: the completeness and detail of the questions and answers in the first round of interrogation, the overlapping content of the two statements, the timing and setting of the first and the second, the continuity of police personnel, and the degree to which the interrogator’s questions treated the second round as continuous with the first. In Elstad, it was not unreasonable to see the occasion for questioning at the station house as presenting a markedly different experience from the short conversation at home; since a reasonable person in the suspect’s shoes could have seen the station house questioning as a new and distinct experience, the Miranda *616warnings could have made sense as presenting a genuine choice whether to follow up on the earlier admission.

At the opposite extreme are the facts here, which by any objective measure reveal a police strategy adapted to undermine the Miranda warnings.6 The unwarned interrogation was conducted in the station house, and the questioning was systematic, exhaustive, and managed with psychological skill. When the police were finished there was little, if anything, of incriminating potential left unsaid. The warned phase of questioning proceeded after a pause of only 15 to 20 minutes, in the same place as the unwarned segment. When the same officer who had conducted the first phase recited the Miranda warnings, he said nothing to counter the probable misimpression that the advice that anything Seibert said could be used against her also applied to the details of the inculpatory, statement previously elicited. In particular, the police did not advise that her prior statement could not be used.7 Nothing was said or done to dispel the oddity of warning about legal rights to silence and counsel right after the police had led her through a systematic interrogation, and any uncertainty on her part about a right to stop talking about matters previously discussed would only have been aggravated by the way Officer Hanrahan set the scene by saying “we’ve been talking for a little while about what happened on Wednesday the twelfth, haven’t we?” App. 66. The impression that the further questioning was a mere continuation of the earlier questions and responses was fostered by references back to the confession already given. It *617would have been reasonable to regard the two sessions as parts of a continuum, in which it would have been unnatural to refuse to repeat at the second stage what had been said before. These circumstances must be seen as challenging the comprehensibility and efficacy of the Miranda warnings to the point that a reasonable person in the suspect’s shoes would not have understood them to convey a message that she retained a choice about continuing to talk.8

VI

Strategists dedicated to draining the substance out of Miranda cannot accomplish by training instructions what Dickerson held Congress could not do by statute. Because the question-first tactic effectively threatens to thwart Miranda’s, purpose of reducing the risk that a coerced confession would be admitted, and because the facts here do not reasonably support a conclusion that the warnings given could have served their purpose, Seibert’s postwarning statements are inadmissible. The judgment of the Supreme Court of Missouri is affirmed.

It is so ordered.

Justice Breyer,

concurring.

In my view, the following simple rule should apply to the two-stage interrogation technique: Courts should exclude the “fruits” of the initial unwarned questioning unless the failure to warn was in good faith. Cf. Oregon v. Elstad, 470 U. S. 298, 309, 318, n. 5 (1985); United States v. Leon, 468 U. S. 897 (1984). I believe this is a sound and workable approach to the problem this case presents. Prosecutors and judges have long understood how to apply the “fruits” approach, which they use in other areas of law. See Wong Sun v. United States, 371 U. S. 471 (1963). And in the workaday *618world of criminal law enforcement the administrative simplicity of the familiar has significant advantages over a more complex exclusionary rule. Cf. post, at 628-629 (O’Connor, J., dissenting).

I believe the plurality’s approach in practice will function as a “fruits” test. The truly “effective” Miranda warnings on which the plurality insists, ante, at 615, will occur only when certain circumstances — a lapse in time, a change in location or interrogating officer, or a shift in the focus of the questioning — intervene between the unwarned questioning and any postwarning statement. Cf. Taylor v. Alabama, 457 U. S. 687, 690 (1982) (evidence obtained subsequent to a constitutional violation must be suppressed as “fruit of the poisonous tree” unless “intervening events break the causal connection”).

I consequently join the plurality’s opinion in full. I also agree with Justice Kennedy’s opinion insofar as it is consistent with this approach and makes clear that a good-faith exception applies. See post, at 622 (opinion concurring in judgment).

Justice Kennedy,

concurring in the judgment.

The interrogation technique used in this case is designed to circumvent Miranda v. Arizona, 384 U. S. 436 (1966). It undermines the Miranda wárning and obscures its meaning. The plurality opinion is correct to conclude that statements obtained through the use of this technique are inadmissible. Although I agree with much in the careful and convincing opinion for the plurality, my approach does differ in some respects, requiring this separate statement.

The Miranda rule has become an important and accepted element of the criminal justice system. See Dickerson v. United States, 530 U. S. 428 (2000). At the same time, not every violation of the rule requires suppression of the evidence obtained. Evidence is admissible when the central *619concerns of Miranda are not likely to be implicated and when other objectives of the criminal justice system are best served by its introduction. Thus, we have held that statements obtained in violation of the rule can be used for impeachment, so that the truth-finding function of the trial is not distorted by the defense, see Harris v. New York, 401 U. S. 222 (1971); that there is an exception to protect countervailing concerns of public safety, see New York v. Quarles, 467 U. S. 649 (1984); and that physical evidence obtained in reliance on statements taken in violation of the rule is admissible, see United States v. Patane, post, p. 630. These cases, in my view, are correct. They recognize that admission of evidence is proper when it would further important objectives without compromising Miranda’s central concerns. Under these precedents, the scope of the Miranda suppression remedy depends on a consideration of those legitimate interests and on whether admission of the evidence under the circumstances would frustrate Miranda’s central concerns and objectives.

Oregon v. Elstad, 470 U. S. 298 (1985), reflects this approach. In Elstad, a suspect made an initial incriminating statement at his home. The suspect had not received a Miranda warning before making the statement, apparently because it was not clear whether the suspect was in custody at the time. The suspect was taken to the station house, where he received a proper warning, waived his Miranda rights, and made a second statement. He later argued that the postwarning statement should be suppressed because it was related to the unwarned first statement, and likely induced or caused by it. The Court held that, although a Miranda violation made the first statement inadmissible, the postwarning statements could be introduced against the accused because “neither the general goal of deterring improper police conduct nor the Fifth Amendment goal of assuring trustworthy evidence would be served by suppres*620sion” given the facts of that case. Elstad, supra, at 308 (citing Michigan v. Tucker, 417 U. S. 433, 445 (1974)).

In my view, Elstad was correct in its reasoning and its result. Elstad reflects a balanced and pragmatic approach to enforcement of the Miranda warning. An officer may not realize that a suspect is in custody and warnings are required. The officer may not plan to question the suspect or may be waiting for a more appropriate time. Skilled investigators often interview suspects multiple times, and good police work may involve referring to prior statements to test their veracity or to refresh recollection. In light of these realities it would be extravagant to treat the presence of one statement that cannot be admitted under Miranda as sufficient reason to prohibit subsequent statements preceded by a proper warning. See Elstad, 470 U. S., at 309 (“It is an unwarranted extension of Miranda to hold that a simple failure to administer the warnings... so taints the investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period”). That approach would serve “neither the general goal of deterring improper police conduct nor the Fifth Amendment goal of assuring trustworthy evidence would be served by suppression of the . . . testimony.” Id., at 308.

This case presents different considerations. The police used a two-step questioning technique based on a deliberate violation of Miranda. The Miranda warning was withheld to obscure both the practical and legal significance of the admonition when finally given. As Justice Souter points out, the two-step technique permits the accused to conclude that the right not to respond did not exist when the earlier incriminating statements were made. The strategy is based on the assumption that Miranda warnings will tend to mean less when recited midinterrogation, after inculpatory statements have already been obtained. This tactic relies on an intentional misrepresentation of the protection that Mi *621 randa offers and does not serve any legitimate objectives that might otherwise justify its use.

Further, the interrogating officer here relied on the defendant’s prewarning statement to obtain the postwarning statement used against her at trial. The postwarning interview resembled a cross-examination. The officer confronted the defendant with her inadmissible prewarning statements and pushed her to acknowledge them. See App. 70 (“ ’Trice, didn’t you tell me that he was supposed to die in his sleep?”). This shows the temptations for abuse inherent in the two-step technique. Reference to the prewarning statement was an implicit suggestion that the. mere repetition of the earlier statement was not independently incriminating. The implicit suggestion was false.

The technique used in this case ’distorts the meaning of Miranda and furthers no legitimate countervailing interest. The Miranda rule would be frustrated were we to allow police to undermine its meaning and effect. The technique simply creates too high a risk that postwarning statements will be obtained when a suspect was deprived of “knowledge essential to his ability to understand the nature of his rights and the consequences of abandoning them.” Moran v. Bur-bine, 475 U. S. 412, 423-424 (1986). When an interrogator uses this deliberate, two-step strategy, predicated upon violating Miranda during an extended interview, postwarning statements that are related to the substance of prewarning statements must be excluded absent specific, curative steps.

The plurality concludes that whenever a two-stage interview occurs, admissibility of the postwarning statement should depend on “whether [the] Miranda warnings delivered midstream could have been effective enough to accomplish their object” given the specific facts of the case. Ante, • at 615. This test envisions an objective inquiry from the perspective of the suspect, and applies in the case of both intentional and unintentional two-stage interrogations. *622 Ante, at 615-617. In my view, this test cuts too broadly. Miranda's clarity is one of its strengths, and a multifactor test that applies to every two-stage interrogation may serve to undermine that clarity. Cf. Berkemer v. McCarty, 468 U. S. 420, 430 (1984). I would apply a narrower test applicable only in the infrequent case, such as we have here, in which the two-step interrogation technique was used in a calculated way to undermine the Miranda warning.

The admissibility of postwarning statements should continue to be governed by the principles of Elstad unless the deliberate two-step strategy was employed. If the deliberate two-step strategy has been used, postwarning statements that are related to the substance of prewarning statements must be excluded unless curative measures are taken before the postwarning statement is made. Curative measures should be designed to ensure that a reasonable person in the suspect’s situation would understand the import and effect of the Miranda warning and of the Miranda waiver. For example, a substantial break in time and circumstances between the prewarning statement and the Miranda warning may suffice in most circumstances, as it allows the accused to distinguish the two contexts and appreciate that the interrogation has taken a new turn. Cf. Westover v. United States, decided with Miranda v. Arizona, 384 U. S. 436 (1966). Alternatively, an additional warning that explains the likely inadmissibility of the prewarning custodial statement may be sufficient. No curative steps were taken in this ease, however, so the postwarning statements are inadmissible and the conviction cannot stand.

For these reasons, I concur in the judgment of the Court.

Justice O’Connor,

with whom The Chief Justice, Justice Scalia, and Justice Thomas join, dissenting.

The plurality devours Oregon v. Elstad, 470 U. S. 298 (1985), even as it accuses petitioner’s argument of “disfiguring]” that decision. Ante, at 614. I believe that we *623are bound by Elstad to reach a different result, and I would vacate the judgment of the Supreme Court of Missouri.

I

On two preliminary questions I am in full agreement with the plurality. First, the plurality appropriately follows El-stad in concluding that Seibert’s statement cannot be held inadmissible under a “fruit of the poisonous tree” theory. Ante, at 612, n. 4 (internal quotation marks omitted). Second, the plurality correctly declines to focus its analysis on the subjective intent of the interrogating officer.

A

This Court has made clear that there simply is no place for a robust deterrence doctrine with regard to violations of Miranda v. Arizona, 384 U. S. 436 (1966). See Dickerson v. United States, 530 U. S. 428, 441 (2000) (“Our decision in [Elstad] — refusing to apply the traditional ‘fruits’ doctrine developed in Fourth Amendment cases — . . . simply recognizes the fact that unreasonable searches under the Fourth Amendment are different from unwarned interrogation under the Fifth Amendment”); Elstad, supra, at 306 (unlike the Fourth Amendment exclusionary rule, the “Miranda exclusionary rule ... serves the Fifth Amendment and sweeps more broadly than the Fifth Amendment itself”); see also United States v. Patane, post, at 644-645 (Kennedy, J., concurring in judgment) (refusal to suppress evidence obtained following an unwarned confession in Elstad, New York v. Quarles, 467 U. S. 649 (1984), and Harris v. New York, 401 U. S. 222 (1971), was based on “our recognition that the concerns underlying the Miranda . . . rule must be accommodated to other objectives of the criminal justice system”). Consistent with that view, the Court today refuses to apply the traditional “fruits” analysis to the physical fruit of a claimed Miranda violation. Patane, post, p. 630. The plu*624rality correctly refuses to apply a similar analysis to testimonial fruits.

Although the analysis the plurality ultimately espouses examines the same facts and circumstances that a “fruits” analysis would consider (such as the lapse of time between the two interrogations and change of questioner or location), it does so for entirely different reasons. The fruits analysis would examine those factors because they are relevant to the balance of deterrence value versus the “drastic and socially costly course” of excluding reliable evidence. Nix v. Williams, 467 U. S. 431, 442-443 (1984). The plurality, by contrast, looks to those factors to inform the psychological judgment regarding whether the suspect has been informed effectively of her right to remain silent. The analytical underpinnings of the two approaches are thus entirely distinct, and they should not be conflated just because they function similarly in practice. Cf. ante, at 617-618 (Breyer, J., concurring).

B

The plurality’s rejection of an intent-based test is also, in my view, correct. Freedom from compulsion lies at the heart of the Fifth Amendment, and requires us to assess whether a suspect’s decision to speak truly was voluntary. Because voluntariness is a matter of the suspect’s state of mind, we focus our analysis on the way in which suspects experience interrogation. See generally Miranda, 384 U. S., at 455 (summarizing psychological tactics used by police that “undermin[e]” the suspect’s “will to resist,” and noting that “the very fact of custodial interrogation . . . trades on the weakness of individuals”); id., at 467 (“[I]n-custody interrogation of persons suspected or accused of crime contains inherently compelling pressures which work to undermine the individual’s will to resist and to compel him to speak where he would not otherwise do so freely”).

Thoughts kept inside a police officer’s head cannot affect that experience. See Moran v. Burbine, 475 U. S. 412, 422 *625(1986) (“Events occurring outside of the presence of the suspect and entirely unknown to him surely can have no bearing on the capacity to comprehend and knowingly relinquish a constitutional right”). In Moran, an attorney hired by the suspect’s sister had been trying to contact the suspect and was told by the police, falsely, that they would not begin an interrogation that night. Id., at 416-418. The suspect was not aware that an attorney had been hired for him. Id., at 417. We rejected an analysis under which a different result would obtain for “the same defendant, armed with the same information and confronted with precisely the same police conduct” if something not known to the defendant — such as the fact that an attorney was attempting to contact him— had been different. Id., at 422. The same principle applies here. A suspect who experienced exactly the same interrogation as Seibert, save for a difference in the undivulged, subjective intent of the interrogating officer when he failed to give Miranda warnings, would not experience the interrogation any differently. “[Wjhether intentional or inadvertent, the state of mind of the police is irrelevant to the question of the intelligence and voluntariness of respondent’s election to abandon his rights. Although highly inappropriate, even deliberate deception of an attorney could not possibly affect a suspect’s decision to waive his Miranda rights unless he were at least aware of the incident.” 475 U. S., at 423. Cf. Stansbury v. California, 511 U. S. 318, 324-325 (1994) (per curiam) (police officer’s subjective intent is irrelevant to whether suspect is in custody for Miranda purposes; “one cannot expect the person under interrogation to probe the officer’s innermost thoughts”).

Because the isolated fact of Officer Hanrahan’s intent could not have had any bearing on Seibert’s “capacity to comprehend and knowingly relinquish” her right to remain silent, Moran, supra, at 422, it could not by itself affect the volun-tariness of her confession. Moreover, recognizing an exception to Elstad for intentional violations would require focus*626ing constitutional analysis on a police officer’s subjective intent, an unattractive proposition that we all but uniformly avoid. In general, “we believe that ‘sending state and federal courts on an expedition into the minds of police officers would produce a grave and fruitless misallocation of judicial resources.’ ” United States v. Leon, 468 U. S. 897, 922, n. 23 (1984) (quoting Massachusetts v. Painten, 389 U. S. 560, 565 (1968) (White, J., dissenting)). This case presents the uncommonly straightforward circumstance of an officer openly admitting that the violation was intentional. But the inquiry will be complicated in other situations probably more likely to occur. For example, different officers involved in an interrogation might claim different states of mind regarding the failure to give Miranda warnings. Even in the simple case of a single officer who claims that a failure to give Miranda warnings was inadvertent, the likelihood of error will be high. See W. LaFave, Search and Seizure § 1.4(e), p. 124 (3d ed. 1996) (“[T]here is no reason to believe that courts can with any degree of success determine in which instances the police had an ulterior motive”).

These evidentiary difficulties have led us to reject an intent-based test in several criminal procedure contexts. For example, in New York v. Quarles, one of the factors that led us to reject an inquiry into the subjective intent of the police officer in crafting a test for the “public safety” exception to Miranda was that officers’ motives will be “largely unverifiable.” 467 U. S., at 656. Similarly, our opinion in Whren v. United States, 517 U. S. 806, 813-814 (1996), made clear that “the evidentiary difficulty of establishing subjective intent” was one of the reasons (albeit not the principal one) for refusing to consider intent in Fourth Amendment challenges generally.

For these reasons, I believe that the approach espoused by Justice Kennedy is ill advised. Justice Kennedy would extend Miranda’s exclusionary rule to any case in which the use of the “two-step interrogation technique” was “deliber*627ate” or “calculated.” Ante, at 622 (opinion concurring in judgment). This approach untethers the analysis from facts knowable to, and therefore having any potential directly to affect, the suspect. Far from promoting “clarity,” ibid., the approach will add a third step to the suppression inquiry. In virtually every two-stage interrogation case, in addition to addressing the standard Miranda and voluntariness questions, courts will be forced to conduct the kind of difficult, state-of-mind inquiry that we normally take pains to avoid.

II

The plurality’s adherence to Elstad, and mine to the plurality, end there. Our decision in Elstad rejected, two lines of argument advanced in favor of suppression. The first was based on the “fruit of the poisonous tree” doctrine, discussed above. The second was the argument that the “lingering compulsion” inherent in a defendant’s having let the “cat out of the bag” required suppression. 470 U. S., at 311. The Court of Appeals of Oregon, in accepting the latter argument, had endorsed a theory indistinguishable from the one today’s plurality adopts: “[T]he coercive impact of the unconstitutionally obtained statement remains, because in a defendant’s mind it has sealed his fate. It is this impact that must be dissipated in order to make a subsequent confession admissible.” State v. Elstad, 61 Ore. App. 673, 677, 658 P. 2d 552, 554 (1983).

We rejected this theory outright. We did so not because we refused to recognize the “psychological impact of the suspect’s conviction that he has let the cat out of the bag,” but because we refused to “endo[w]” those “psychological effects” with “constitutional implications.” 470 U. S., at 311. To do so, we said, would “effectively immuniz[e] a suspect who responds to pre-Miranda warning questions from the consequences of his subsequent informed waiver,” an immunity that “comes at a high cost to legitimate law enforcement activity, while adding little desirable protection to the indi*628vidual’s interest in not being compelled to testify against himself.” Id., at 312. The plurality might very well think that we struck the balance between Fifth Amendment rights and law enforcement interests incorrectly in Elstad; but that is not normally a sufficient reason for ignoring the dictates of stare decisis.

I would analyze the two-step interrogation procedure under the voluntariness standards central to the Fifth. Amendment and reiterated in Elstad. Elstad commands that if Seibert’s first statement is shown to have been involuntary, the court must examine whether the taint dissipated through the passing of time or a change in circumstances: “When a prior statement is actually coerced, the time that passes between confessions, the change in place of interrogations, and the change in identity of the interrogators all bear on whether that coercion has carried over into the second confession.” Id., at 310 (citing Westover v. United States, decided with Miranda, 384 U. S., at 494). In addition, ,Sei-bert’s second statement should be suppressed if she showed that it was involuntary despite the Miranda warnings. El-stad, supra, at 318 (“The relevant inquiry is whether, in fact, the second statement was also voluntarily made. As in any such inquiry, the finder of fact must examine the surrounding circumstances and the entire course of police conduct with respect to the suspect in evaluating the voluntariness of his statements”). Although I would leave this analysis for the Missouri courts to conduct on remand, I note that, unlike the officers in Elstad, Officer Hanrahan referred to Seibert’s unwarned statement during the second part of the interrogation when she made a statement at odds with her unwarned confession. App. 70 (“ ’Trice, didn’t you tell me that he was supposed to die in his sleep?”); cf. Elstad, supra, at 316 (officers did not “exploit the unwarned admission to pressure respondent into waiving his right to remain silent”). Such a tactic may bear on the voluntariness inquiry. Cf. Frazier v. Cupp, 394 U. S. 731, 739 (1969) (fact that police had falsely *629told a suspect that his accomplice had already confessed was “relevant” to the voluntariness inquiry); Moran, 475 U. S., at 423-424 (in discussing police deception, stating that simply withholding information is “relevant to the constitutional validity of a waiver if it deprives a defendant of knowledge essential to his ability to understand the nature of his rights and the consequences of abandoning them”); Miranda, supra, at 476.

* * *

Because I believe that the plurality gives insufficient deference to Elstad and that Justice Kennedy places improper weight on subjective intent, I respectfully dissent.

Mr. Chief Justice Burger

delivered the opinion of the Court.

We granted the writ in this case to consider petitioner’s claim that a statement made by him to police under circumstances rendering it inadmissible to establish the prosecution’s case in chief under Miranda v. Arizona, 384 U. S. 436 (1966), may not be used to impeach his credibility.

The State of New York charged petitioner in a two-count indictment with twice selling heroin to an under*223cover police officer. At a subsequent jury trial the officer was the State’s chief witness, and he testified as to details of the two sales. A second officer verified collateral details of the sales, and a third offered testimony about the chemical analysis of the heroin.

Petitioner took the stand in his own defense. He admitted knowing the undercover police officer but denied a sale on January 4, 1966. He admitted making a sale of contents of a glassine bag to the officer on January 6 but claimed it was baking powder and part of a scheme to defraud the purchaser.

On cross-examination petitioner was asked seriatim whether he had made specified statements to the police immediately following his arrest on January 7 — statements that partially contradicted petitioner’s direct testimony at trial. In response to the cross-examination, petitioner testified that he could not remember virtually any of the questions or answers recited by the prosecutor. At the request of petitioner’s counsel the written statement from which the prosecutor had read questions and answers in his impeaching process was placed in the record for possible use on appeal; the statement was not shown to the jury.

The trial judge instructed the jury that the statements attributed to petitioner by the prosecution could be considered only in passing on petitioner’s credibility and not as evidence of guilt. In closing summations both counsel argued the substance of the impeaching statements. The jury then found petitioner guilty on the second count of the indictment.1 The New York Court of Appeals affirmed in a per curiam opinion, 25 N. Y. 2d 175, 250 N. E. 2d 349 (1969).

At trial the prosecution made no effort in its case in chief to use the statements allegedly made by petitioner, *224conceding that they were inadmissible under Miranda v. Arizona, 384 U. S. 436 (1966). The transcript of the interrogation used in the impeachment, but not given to the jury, shows that no warning of a right to appointed counsel was given before questions were put to petitioner when he was taken into custody. Petitioner makes no claim that the statements made to the police were coerced or involuntary.

Some comments in the Miranda opinion can indeed be read as indicating a bar to use of an uncounseled statement for any purpose, but discussion of that issue was not at all necessary to the Court’s holding and cannot be regarded as controlling. Miranda barred the prosecution from making its case with statements of an accused made while in custody prior to having or effectively waiving counsel. It does not follow from Miranda that evidence inadmissible against an accused in the prosecution’s case in chief is barred for all purposes, provided of course that the trustworthiness of the evidence satisfies legal standards.

In Walder v. United States, 347 U. S. 62 (1954), the Court permitted physical evidence, inadmissible in the case in chief, to be used for impeachment purposes.

“It is one thing to say that the Government cannot make an affirmative use of evidence unlawfully obtained. It is quite another to say that the defendant can turn the illegal method by which evidence in the Government’s possession was obtained to his own advantage, and provide himself with a shield against contradiction of his untruths. Such an extension of the Weeks doctrine would be a perversion of the Fourth Amendment.

“[T]here is hardly justification for letting the defendant affirmatively resort to perjurious testimony in reliance on the Government’s disability to challenge his credibility.” 347 U. S., at 65.

*225It is true that Walder was impeached as to collateral matters included in his direct examination, whereas petitioner here was impeached as to testimony bearing more directly on the crimes charged. We are not persuaded that there is a difference in principle that warrants a result different from that reached by the Court in Walder. Petitioner's testimony in his own behalf concerning the events of January 7 contrasted sharply with what he told the police shortly after his arrest. The impeachment process here undoubtedly provided valuable aid to the jury in assessing petitioner’s credibility, and the benefits of this process should not be lost, in our view, because of the speculative possibility that impermissible police conduct will be encouraged thereby. Assuming that the exclusionary rule has a deterrent effect on proscribed police conduct, sufficient deterrence flows when the evidence in question is made unavailable to the prosecution in its case in chief.

Every criminal defendant is privileged to testify in his own defense, or to refuse to do so. But that privilege cannot be construed to include the right to commit perjury. See United States v. Knox, 396 U. S. 77 (1969); cf. Dennis v. United States, 384 U. S. 855 (1966). Having voluntarily taken the stand, petitioner was under an obligation to speak truthfully and accurately, and the prosecution here did no more than utilize the traditional truth-testing devices of the adversary process.2 Had *226inconsistent statements been made by the accused to some third person, it could hardly be contended that the conflict could not be laid before the jury by way of cross-examination and impeachment.

The shield provided by Miranda cannot be perverted into a license to use perjury by way of a defense, free from the risk of confrontation with prior inconsistent utterances. We hold, therefore, that petitioner’s credibility was appropriately impeached by use of his earlier conflicting statements.

Affirmed.

Me. Justice Black dissents.

Mr. Justice Brennan,

with whom Mr. Justice Douglas and Mr. Justice Marshall join, dissenting.

It is conceded that the question-and-answer statement used to impeach petitioner’s direct testimony was, under Miranda v. Arizona, 384 U. S. 436 (1966), constitutionally inadmissible as part of the State’s direct case against petitioner. I think that the Constitution also denied the State the use of the statement on cross-examination to impeach the credibility of petitioner’s testimony given in his own defense. The decision in Walder v. United States, 347 U. S. 62 (1954), is not, as the Court today holds, dispositive to the contrary. Rather, that case supports my conclusion.

The State’s case against Harris depended upon the jury’s belief of the testimony of the undercover agent that petitioner “sold” the officer heroin on January 4 and again on January 6. Petitioner took the stand and flatly denied having sold anything to the officer on January 4. He countered the officer’s testimony as to the January 6 sale with testimony that he had sold the officer two glassine bags containing what appeared to be heroin, but that actually the bags contained only baking powder intended to deceive the officer in order to obtain $12. *227The statement contradicted petitioner’s direct testimony as to the events of both days. The statement’s version of the events on January 4 was that the officer had used petitioner as a middleman to buy some heroin from a third person with money furnished by the officer. The version of the events on January 6 was that petitioner had again acted for the officer in buying two bags of heroin from a third person for which petitioner received $12 and a part of the heroin. Thus, it is clear that the statement was used to impeach petitioner’s direct testimony not on collateral matters but on matters directly related to the crimes for which he was on trial.1

Walder v. United States was not a case where tainted evidence was used to impeach an accused’s direct testimony on matters directly related to the case against him. In Walder the evidence was used to impeach the accused’s testimony on matters collateral to the crime charged. Walder had been indicted in 1950 for purchasing and possessing heroin. When his motion to suppress use of the narcotics as illegally seized was granted, the Government dismissed the prosecution. Two years later Walder was indicted for another narcotics violation completely unrelated to the 1950 one. Testifying in his own defense, he said on direct examination that he had never in his life possessed narcotics. On cross-examination he denied that law enforcement officers had seized narcotics from his home two years earlier. The Government was then permitted to introduce the testimony of one of the officers involved in the 1950 seizure, that when he had raided Walder’s home at that time he had seized narcotics there. *228The Court held that on facts where “the defendant went beyond a mere denial of complicity in the crimes of which he was charged and made the sweeping claim that he had never dealt in or possessed any narcotics,” 347 U. S., at 65, the exclusionary rule of Weeks v. United States, 232 U. S. 383 (1914), would not extend to bar the Government from rebutting this testimony with evidence, although tainted, that petitioner had in fact possessed narcotics two years before. The Court was careful, however, to distinguish the situation of an accused whose testimony, as in the instant case, was- a “denial of complicity in the crimes of which he was charged,” that is, where illegally obtained evidence was used to impeach the accused’s direct testimony on matters directly related to the case against him. As to that situation, the Court said:

“Of course, the Constitution guarantees a defendant the fullest opportunity to meet the accusation against him. He must be free to deny all the elements of the case against him without thereby giving leave to the Government to introduce by way of rebuttal evidence illegally secured by it, and therefore not available for its case in chief.” 347 U. S., at 65.

From this recital of facts it is clear that the evidence used for impeachment in Walder was related to the earlier 1950 prosecution and had no direct bearing on “the elements of the case” being tried in 1952. The evidence tended solely to impeach the credibility of the defendant’s direct testimony that he had never in his life possessed heroin. But that evidence was completely unrelated to the indictment on trial and did not in any way interfere with his freedom to deny all elements of that case against him. In contrast, here, the evidence used for impeachment, a statement concerning the details of the very sales alleged in the indictment, was directly related to the case against petitioner.

*229While Walder did not identify the constitutional specifics that guarantee “a defendant the fullest opportunity to meet the accusation against him . . . [and permit him to] be free to deny all the elements of the case against him,” in my view Miranda v. Arizona, 384 U. S. 436 (1966), identified the Fifth Amendment’s privilege against self-incrimination as one of those specifics.2 *230That privilege has been extended against the States. Malloy v. Hogan, 378 U. S. 1 (1964). It is fulfilled only when an accused is guaranteed the right “to remain silent unless he chooses to speak in the unfettered exercise of his own will,” id., at 8 (emphasis added). The choice of whether to testify in one’s own defense must therefore be “unfettered,” since that choice is an exercise of the constitutional privilege, Griffin v. California, 380 U. S. 609 (1965). Griffin held that comment by the prosecution upon the accused’s failure to take the stand or a court instruction that such silence is evidence of guilt is impermissible because it “fetters” that choice — “[i]t cuts down on the privilege by making its assertion costly.” Id., at 614. For precisely the same reason the constitutional guarantee forbids the prosecution to use a tainted statement to impeach the accused who takes the stand: The prosecution’s use of the tainted statement “cuts down on the privilege by making its assertion costly.” Ibid. Thus, the accused is denied an “unfettered” choice when the decision whether to take the stand is burdened by the risk that an illegally obtained prior statement may be introduced to impeach his direct testimony denying complicity in the crime charged against him.3 We settled this proposition in Miranda where we said:

“The privilege against self-incrimination protects the individual from being compelled to incriminate himself in any manner .... [Statements merely intended to be exculpatory by the defendant are often used to impeach his testimony at trial . . . . These statements are incriminating in any meaningful sense of the word and may not be used without the full warnings and effective waiver required for *231 any other statement.” 384 U. S., at 476-477 (emphasis added).

This language completely disposes of any distinction between statements used on direct as opposed to cross-examination.4 “An incriminating statement is as incriminating when used to impeach credibility as it is when used as direct proof of guilt and no constitutional distinction can legitimately be drawn.” People v. Kulis, 18 N. Y. 2d 318, 324, 221 N. E. 2d 541, 543 (1966.) (dissenting opinion).

The objective of deterring improper police conduct is only part of the larger objective of safeguarding the integrity of our adversary system. The “essential mainstay” of that system, Miranda v. Arizona, 384 U. S., at 460, is the privilege against self-incrimination, which for *232that reason has occupied a central place in our jurisprudence since before the Nation’s birth. Moreover, “we may view the historical development of the privilege as one which groped for the proper scope of governmental power over the citizen. . . . All these policies point to one overriding thought: the constitutional foundation underlying the privilege is the respect a government . . . must accord to the dignity and integrity of its citizens.” Ibid. These values are plainly jeopardized if an exception against admission of tainted statements is made for those used for impeachment purposes. Moreover, it is monstrous that courts should aid or abet the law-breaking police officer. It is abiding truth that “[njothing can destroy a government more quickly than its failure to observe its own laws, or worse, its disregard of the charter of its own existence.” Mapp v. Ohio, 367 U. S. 643, 659 (1961). Thus, even to the extent that Miranda was aimed at deterring police practices in disregard of the Constitution, I fear that today’s holding will seriously undermine the achievement of that objective. The Court today tells the police that they may freely interrogate an accused incommunicado and without counsel and know that although any statement they obtain in violation of Miranda cannot be used on the State’s direct case, it may be introduced if the defendant has the temerity to testify in his own defense. This goes far toward undoing much of the progress made in conforming police methods to the Constitution. I dissent.

Justice Brennan

delivered the opinion of the Court.

The impeachment exception to the exclusionary rule permits the prosecution in a criminal proceeding to introduce il*309legally obtained evidence to impeach the defendant’s own testimony. The Illinois Supreme Court extended this exception to permit the prosecution to impeach the testimony of all defense witnesses with illegally obtained evidence. 123 Ill. 2d 523, 528 N. E. 2d 723 (1988). Finding this extension inconsistent with the balance of values underlying our previous applications of the exclusionary rule, we reverse.

H-l

On the night of August 30, 1982, eight young boys returning home from a party were confronted by a trio of other boys who demanded money. When the eight boys refused to comply, one member of the trio produced a gun and fired into the larger group, killing one boy and seriously injuring another. When the police arrived, the remaining members of the larger group provided eyewitness accounts of the event and descriptions of the perpetrators.

The next evening, two detectives of the Chicago Police Department took 15-year-old Darryl James into custody as a suspect in the shooting. James was found at his mother’s beauty parlor sitting under a hair dryer; when he emerged, his hair was black and curly. After placing James in their car, the detectives questioned him about his prior hair color. He responded that the previous day his hair had been reddish brown, long, and combed straight back. The detectives questioned James again later at the police station, and he further stated that he had gone to the beauty parlor in order to have his hair “dyed black and curled in order to change his appearance.” App. 11.

The State subsequently indicted James for murder and attempted murder. Prior to trial, James moved to suppress the statements regarding his hair, contending that they were the fruit of a Fourth Amendment violation because the detectives lacked probable cause for his warrantless arrest. After an evidentiary hearing, the trial court sustained this *310motion and ruled that the statements would be inadmissible at trial.

At trial, five members of the larger group of boys testified for the State, and each made an in-court identification of the defendant. Each testified that the person responsible for the shooting had “reddish” hair, worn shoulder length in a slicked-back “butter” style. Each also recalled having seen James several weeks earlier at a parade, at which time James had the aforementioned hair color and style. At trial, however, his hair was black and worn in a “natural” style. Despite the discrepancy between the witnesses’ description and his present appearance, the witnesses stood firm in their conviction that James had been present and had fired the shots.

James did not testify in his own defense. He called as a witness Jewel Henderson, a friend of his family. Henderson testified that on the day of the shooting she had taken James to register for high school and that, at that time, his hair was black. The State then sought, over James’ objection, to introduce his illegally obtained statements as a means of impeaching the credibility of Henderson’s testimony. After determining that the suppressed statements had been made voluntarily, the trial court overruled James’ objection. One of the interrogating detectives then reported James’ prior admissions that he had reddish hair the night of the shooting and he dyed and curled his hair the next day in order to change his appearance. James ultimately was convicted of both murder and attempted murder and sentenced to 30 years’ imprisonment.

On appeal, the Illinois Appellate Court reversed James’ convictions and ordered a new trial. 153 Ill. App. 3d 131, 505 N. E. 2d 1118 (1987). The appellate court held that the exclusionary rule barred admission of James’ illegally obtained statements for the purpose of impeaching a defense witness’ testimony and that the resulting constitutional error was not harmless. However, the Illinois Supreme Court re*311versed. The court reasoned that, in order to deter the defendant from engaging in perjury “by proxy,” the impeachment exception to the exclusionary rule ought to be expanded to allow the State to introduce illegally obtained evidence to impeach the testimony of defense witnesses other than the defendant himself. The court therefore ordered James’ convictions reinstated. We granted certiorari. 489 U. S. 1010 (1989).

II

“There is no gainsaying that arriving at the truth is a fundamental goal of our legal system.” United States v. Havens, 446 U. S. 620, 626 (1980). But various constitutional rules limit the means by which government may conduct this search for truth in order to promote other values embraced by the Framers and cherished throughout our Nation’s history. “Ever since its inception, the rule excluding evidence seized in violation of the Fourth Amendment has been recognized as a principal mode of discouraging lawless police conduct. . . . [WJithout it the constitutional guarantee against unreasonable searches and seizures would be a mere ‘form of words.’” Terry v. Ohio, 392 U. S. 1, 12 (1968), quoting Mapp v. Ohio, 367 U. S. 643, 655 (1961). The occasional suppression of illegally obtained yet probative evidence has long been considered a necessary cost of preserving overriding constitutional values: “[T]here is nothing new in the realization that the Constitution sometimes insulates the criminality of a few in order to protect the privacy of us all.” Arizona v. Hicks, 480 U. S. 321, 329 (1987).

This Court has carved out exceptions to the exclusionary rule, however, where the introduction of reliable and probative evidence would significantly further the truth-seeking function of a criminal trial and the likelihood that admissibility of such evidence would encourage police misconduct is but a “speculative possibility.” Harris v. New York, 401 U. S. *312222, 225 (1971).1 One exception to the rule permits prosecutors to introduce illegally obtained evidence for the limited purpose of impeaching the credibility of the defendant’s own testimony. This Court first recognized this exception in Walder v. United States, 347 U. S. 62 (1954), permitting the prosecutor to introduce into evidence heroin obtained through an illegal search to undermine the credibility of the defendant’s claim that he had never possessed narcotics. The Court explained that a defendant

“must be free to deny all the elements of the case against him without thereby giving leave to the Government to introduce by way of rebuttal evidence illegally secured by it, and therefore not available for its case in chief. Beyond that, however, there is hardly justification for letting the defendant affirmatively resort to perjurious testimony in reliance on the Government’s disability to challenge his credibility.” Id., at 65.

In Harris v. New York, supra, and Oregon v. Hass, 420 U. S. 714 (1975), the Court applied the exception to permit prosecutors to impeach defendants using incriminating yet voluntary and reliable statements elicited in violation of Miranda requirements.2 Finally, in United States v. Havens, supra, the Court expanded the exception to permit *313prosecutors to introduce illegally obtained evidence in order to impeach a defendant’s “answers to questions put to him on cross-examination that are plainly within the scope of the defendant’s direct examination.” Id., at 627.

This Court insisted throughout this line of cases that “evidence that has been illegally obtained ... is inadmissible on the government’s direct case, or otherwise, as substantive evidence of guilt.” Id., at 628.3 However, because the Court believed that permitting the use of such evidence to impeach defendants’ testimony would further the goal of truthseeking by preventing defendants from perverting the exclusionary rule “ ‘into a license to use perjury by way of a defense,”’ id., at 626 (citation omitted), and because the Court further believed that permitting such use would create only a “speculative possibility that impermissible police conduct will be encouraged thereby,” Harris, supra, at 225, the Court concluded that the balance of values underlying the exclusionary rule justified an exception covering impeachment of defendants’ testimony.

Ill

In this case, the Illinois Supreme Court held that our balancing approach in Walder and its progeny justifies expanding the scope of the impeachment exception to permit prosecutors to use illegally obtained evidence to impeach the credibility of defense witnesses. We disagree. Expanding the class of impeachable witnesses from the defendant alone to all defense witnesses would create different incentives affecting the behavior of both defendants and law enforcement officers. As a result, this expansion would not promote the truth-seeking function to the same extent as did creation of the original exception, and yet it would significantly under*314mine the deterrent effect of the general exclusionary rule. Hence, we believe that this proposed expansion would frustrate rather than further the purposes underlying the exclusionary rule.

The previously recognized exception penalizes defendants for committing perjury by allowing the prosecution to expose their perjury through impeachment using illegally obtained evidence. Thus defendants are discouraged in the first instance from “affirmatively resort[ing] to perjurious testimony.” Walder, supra, at 65. But the exception leaves defendants free to testify truthfully on their own behalf; they can offer probative and exculpatory evidence to the jury without opening the door to impeachment by carefully avoiding any statements that directly contradict the suppressed evidence. The exception thus generally discourages perjured testimony without discouraging truthful testimony.

In contrast, expanding the impeachment exception to encompass the testimony of all defense witnesses would not have the same beneficial effects. First, the mere threat of a subsequent criminal prosecution for perjury is far more likely to deter a witness from intentionally lying on a defendant’s behalf than to deter a defendant, already facing conviction for the underlying offense, from lying on his own behalf. Hence the Illinois Supreme Court’s underlying premise that a defendant frustrated by our previous impeachment exception can easily find a witness to engage in “perjury by proxy” is suspect.4

More significantly, expanding the impeachment exception to encompass the testimony of all defense witnesses likely would chill some defendants from presenting their best de*315fense — and sometimes any defense at all — through the testimony of others. Whenever police obtained evidence illegally, defendants would have to assess prior to trial the likelihood that the evidence would be admitted to impeach the otherwise favorable testimony of any witness they call. Defendants might reasonably fear that one or more of their witnesses, in a position to offer truthful and favorable testimony, would also make some statement in sufficient tension with the tainted evidence to allow the prosecutor to introduce that evidence for impeachment. First, defendants sometimes need to call “reluctant” or “hostile” witnesses to provide reliable and probative exculpatory testimony, and such witnesses likely will not share the defendants’ concern for avoiding statements that invite impeachment through contradictory evidence. Moreover, defendants often cannot trust even “friendly” witnesses to testify without subjecting themselves to impeachment, simply due to insufficient care or attentiveness. This concern is magnified in those occasional situations when defendants must call witnesses to testify despite having had only a limited opportunity to consult with or prepare them in advance. For these reasons, we have recognized in a variety of contexts that a party “cannot be absolutely certain that his witnesses will testify as expected.” Brooks v. Tennessee, 406 U. S. 605, 609 (1972).6 As a re-*316suit, an expanded impeachment exception likely would chill some defendants from calling witnesses who would otherwise offer probative evidence.6

*317This realization alters the balance of values underlying the current impeachment exception governing defendants’ testimony. Our prior cases make clear that defendants ought not be able to “pervert” the exclusion of illegally obtained evidence into a shield for perjury, but it seems no more appropriate for the State to brandish such evidence as a sword with which to dissuade defendants from presenting a meaningful defense through other witnesses. Given the potential chill created by expanding the impeachment exception, the conceded gains to the truth-seeking process from discouraging or disclosing perjured testimony would be offset to some extent by the concomitant loss of probative witness testimony. Thus, the truth-seeking rationale supporting the impeachment of defendants in Walder and its progeny does not apply to other witnesses with equal force.

Moreover, the proposed expansion of the current impeachment exception would significantly weaken the exclusionary rule’s deterrent effect on police misconduct. This Court has characterized as a mere “speculative possibility,” Harris v. New York, 401 U. S., at 225, the likelihood that permitting prosecutors to impeach defendants with illegally obtained *318evidence would encourage police misconduct. Law enforcement officers will think it unlikely that the defendant will first decide to testify at trial and will also open the door inadvertently to admission of any illegally obtained evidence. Hence, the officers’ incentive to acquire evidence through illegal means is quite weak.

In contrast, expanding the impeachment exception to all defense witnesses would significantly enhance the expected value to the prosecution of illegally obtained evidence. First, this expansion would vastly increase the number of occasions on which such evidence could be used. Defense witnesses easily outnumber testifying defendants, both because many defendants do not testify themselves and because many if not most defendants call multiple witnesses on their behalf. Moreover, due to the chilling effect identified above, see supra, at 315-316, illegally obtained evidence holds even greater value to the prosecution for each individual witness than for each defendant. The prosecutor’s access to impeachment evidence would not just deter perjury; it would also deter defendants from calling witnesses in the first place, thereby keeping from the jury much probative exculpatory evidence. For both of these reasons, police officers and their superiors would recognize that obtaining evidence through illegal means stacks the deck heavily in the prosecution’s favor. It is thus far more than a “speculative possibility” that police misconduct will be encouraged by permitting such use of illegally obtained evidence.

The United States argues that this result is constitutionally acceptable because excluding illegally obtained evidence solely from the prosecution’s case in chief would still provide a quantum of deterrence sufficient to protect the privacy interests underlying the exclusionary rule.7 We disagree. Of course, a police officer might in certain situations believe that obtaining particular evidence through illegal means, re-*319suiting in,its suppression from the case in chief, would prevent the prosecution from establishing a prima facie case to take to a jury. In such situations, the officer likely would be deterred from obtaining the evidence illegally for fear of jeopardizing the entire case. But much if not most of the time, police officers confront opportunities to obtain evidence illegally after they have already legally obtained (or know that they have other means of legally obtaining) sufficient evidence to sustain a prima facie case. In these situations, a rule requiring exclusion of illegally obtained evidence from only the government’s case in chief would leave officers with little to lose and much to gain by overstepping constitutional limits on evidence gathering.8 Narrowing the exclusionary rule in this manner, therefore, would significantly undermine the rule’s ability “to compel respect for the constitutional guaranty in the only effectively available way — by removing the incentive to disregard it.” Elkins v. United States, 364 U. S. 206, 217 (1960). So long as we are committed to protecting the people from the disregard of their constitutional rights during the course of criminal investigations, inadmissibility of illegally obtained evidence must remain the rule, not the exception.

IV

The cost to the truth-seeking process of evidentiary exclusion invariably is perceived more tangibly in discrete prosecutions than is the protection of privacy values through deterrence of future police misconduct. When defining the precise scope of the exclusionary rule, however, we must focus on systemic effects of proposed exceptions to ensure *320that individual liberty from arbitrary or oppressive police conduct does not succumb to the inexorable pressure to introduce all incriminating evidence, no matter how obtained, in each and every criminal case. Our previous recognition of an impeachment exception limited to the testimony of defendants reflects a careful weighing of the competing values. Because expanding the exception to encompass the testimony of all defense witnesses would not further the truth-seeking value with equal force but would appreciably undermine the deterrent effect of the exclusionary rule, we adhere to the line drawn in our previous cases.

Accordingly, we hold that the Illinois Supreme Court erred in affirming James’ convictions despite the prosecutor’s use of illegally obtained statements to impeach a defense witness’ testimony. The court’s judgment is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

Justice Stevens,

concurring.

While I join the opinion of the Court, certain comments in the dissent prompt this postscript. The dissent answers the wrong question when it states that “[t]he interest in protecting the truth-seeking function of the criminal trial is every bit as strong in this case as in our earlier cases.” Post, at 324. This is self-evident. The State always has a strong interest in the truth-seeking function. The proper question, however, is whether the admission of the illegally obtained evidence in this case would sufficiently advance the truth-seeking function to overcome the loss to the deterrent value of the exclusionary rule. With respect to this issue, the dissent overestimates the benefit of the exclusionary rule even to the defendant bent on presenting perjured testimony and exaggerates the injury that exclusion of unlawfully obtained evidence causes to the truth-seeking function.

In “contested criminal trials,” post, at 326, the urge to win can unfortunately lead each side to overstate its case. As *321the Court properly observes, the ability of the dishonest defendant to procure false testimony is tempered by the availability of the illegally obtained evidence for use in a subsequent perjury prosecution of the defense witness. Ante, at 314. A witness who is not on trial faces a far different calculus than one whose testimony can mean the difference between acquittal and a prison sentence. He or she will think long and hard before accepting a defendant’s invitation to knowingly offer false testimony that is directly contradicted by the State’s evidence. The dissent ignores this “hard reality,” post, at 326, in presuming that a defense witness will offer false testimony when that testimony is immunized from rebuttal at trial.

While the dissent assumes false testimony or, at least, faulty recollection with respect to defense witnesses, it is unwilling to entertain the same assumption with respect to the prosecution’s witnesses. The evidentiary issue in this case involves the testimony of a police officer about a statement that he allegedly heard the defendant make at the time of his arrest. An officer whose testimony provides the foundation for admission of an oral statement or physical evidence may be influenced by his interest in effective law enforcement or may simply have faulty recollection. It is only by giving 100-percent credence to every word of the officer’s testimony that the dissent can so categorically state that “the defendant himself revealed the witness’ testimony to be false,” post, at 324, that “James. . . said his hair was previously red,” post, at 327, n. 2, or that information presented to the jury was “known to be untrue,” post, at 327. That assumption is no more warranted in the case of prosecution witnesses than the opposite assumption is warranted in the case of defense witnesses.

In this case, in which the guilty verdict is supported by the testimony of five eyewitnesses, it is highly probable that these characterizations are accurate. But the testimony of those five witnesses, on which the dissenters rely for their conclusion that any error committed by the trial court was *322harmless, post, at 380, would also seem to be sufficient to obviate the need to rely on the officer’s rebuttal to discredit the witness Henderson’s testimony. Were the officer’s testimony not so corroborated, it would surely be improper to presume — as the dissenters do — that the conflict between the testimony of the officer and Henderson should necessarily be resolved in the officer’s favor or that exclusion of the evidence would result in a decision by jurors who are “positively misled.” Post, at 324.

Justice Kennedy,

with whom The Chief Justice, Justice O’Connor, and Justice ScALiAjoin, dissenting.

To deprive the prosecution of probative evidence acquired in violation of the law may be a tolerable and necessary cost of the exclusionary rule. Implementation of the rule requires us to draw certain lines to effect its purpose of deterring unlawful conduct. But the line drawn by today’s opinion grants the defense side in a criminal case broad immunity to introduce whatever false testimony it can produce from the mouth of a friendly witness. Unless petitioner’s conviction is reversed, we are told, police would flout the Fourth Amendment, and as a result, the accused would be unable to offer any defense. This exaggerated view leads to a drastic remedy: The jury cannot learn that defense testimony is inconsistent with probative evidence of undoubted value. A more cautious course is available, one that retains Fourth Amendment protections and yet safeguards the truth-seeking function of the criminal trial.

Our precedents establish that the exclusionary rule does not apply where the interest in pursuing truth or other important values outweighs any deterrence of unlawful conduct that the rule might achieve. See, e. g., Illinois v. Krull, 480 U. S. 340, 347-348 (1987); United States v. Leon, 468 U. S. 897, 906-907 (1984); Stone v. Powell, 428 U. S. 465, 486-489 (1976); United States v. Calandra, 414 U. S. 338, 347-348 (1974). One instance is a defendant’s attempt to take advantage by presenting testimony in outright contradiction of ex-*323eluded facts, secure in the knowledge that the inconsistency will not be revealed to the jury. As we said over 35 years ago:

“It is one thing to say that the Government cannot make an affirmative use of evidence unlawfully obtained. It is quite another to say that the defendant can turn the illegal method by which evidence in the Government’s possession was obtained to his own advantage, and provide himself with a shield against contradiction of his untruths. Such an extension of the Weeks [v. United States, 232 U. S. 383 (1914),] doctrine would be a perversion of the Fourth Amendment.” Walder v. United States, 347 U. S. 62, 65 (1954).

Under this rationale, our consistent rule has been that a defendant’s testimony is subject to rebuttal by contradicting evidence that otherwise would be excluded. The principle applies to suppressed physical evidence, as in Walder itself and United States v. Havens, 446 U. S. 620 (1980), and to statements obtained in violation of the law, so long as the statements are voluntary and reliable, see Oregon v. Hass, 420 U. S. 714 (1975); Harris v. New York, 401 U. S. 222 (1971).

Petitioner argues that the rationale of these cases is confined to “impeachment” of testimony presented by the defendant himself because these cases involve only “impeachment by self-contradiction.” Brief for Petitioner 13. The theory, it seems, is that excluded evidence introduced in opposition to the defendant’s testimony impeaches by means of the contradiction itself; the substantive truth or falsity of the suppressed evidence is irrelevant. Our cases do not bear this reading. In Havens, the defendant was charged as an accomplice in the smuggling of narcotics. A codefendant hid the drugs in a T-shirt constructed with special pockets. The pockets were made of patches cut from another T-shirt found in the defendant’s luggage during an illegal search. When the defendant denied having possessed the T-shirts, the cut *324T-shirt, which had been excluded at the outset, was admitted as rebuttal evidence. We upheld its admission. See 446 U. S., at 623, 628. There was no “self-contradiction” involved, for the rebuttal of the defendant’s testimony could only have been based on the jury’s belief in the substantive truth of the fact that the altered T-shirt was used in the smuggling, and that it belonged to the defendant. The same was true in Walder, where we upheld the admission of illegally seized heroin from an unrelated investigation to impeach the defendant’s statement that he had never possessed the drug. In sum, our cases show that introduction of testimony contrary to excluded but reliable evidence subjects the testimony to rebuttal by that evidence.

I agree with the majority that the resolution of this case depends on a balance of values that informs our exclusionary rule jurisprudence. We weigh the “‘likelihood of. . . deterrence against the costs of withholding reliable information from the truth-seeking process.’” Ante, at 312, n. 1 (quoting Illinois v. Krull, supra, at 347). The majority adopts a sweeping rule that the testimony of witnesses other than the defendant may never be rebutted with excludable evidence. I cannot draw the line where the majority does.

The interest in protecting the truth-seeking function of the criminal trial is every bit as strong in this case as in our earlier cases that allowed rebuttal with evidence that was inadmissible as part of the prosecution’s case in chief. Here a witness who knew the accused well took the stand to testify about the accused’s personal appearance. The testimony could be expected to create real doubt in the minds of jurors concerning the eyewitness identifications by persons who did not know the accused. To deprive the jurors of knowledge that statements of the defendant himself revealed the witness’ testimony to be false would result in a decision by triers of fact who were not just kept in the dark as to excluded evidence, but positively misled. The potential for harm to the truth-seeking process resulting from the majority’s new rule *325in fact will be greater than if the defendant himself had testified. It is natural for jurors to be skeptical of self-serving testimony by the defendant. Testimony by a witness said to be independent has the greater potential to deceive. And if a defense witness can present false testimony with impunity, the jurors may find the rest of the prosecution’s case suspect, for ineffective and artificial cross-examination will be viewed as a real weakness in the state’s case. Jurors will assume that if the prosecution had any proof the statement was false, it would make the proof known. The majority does more than deprive the prosecution of evidence. The state must also suffer the introduction of false testimony and appear to bolster the falsehood by its own silence.

The majority’s fear that allowing the jury to know the whole truth will chill defendants from putting on any defense seems to me far too speculative to justify the rule here announced. No restriction on the defense results if rebuttal of testimony by witnesses other than the defendant is confined to the introduction of excludable evidence that is in direct contradiction of the testimony. If mere “tension with the tainted evidence,” ante, at 315, opened the door to introduction of all the evidence subject to suppression, then the majority’s fears might be justified. But in this context rebuttal can and should be confined to situations where there is direct conflict, which is to say where, within reason, the witness’ testimony and the excluded testimony cannot both be true.1

*326Also missing from the majority’s analysis is the almost certain knowledge that the testimony immunized from rebuttal is false. The majority’s apparent assumption that defense witnesses protected by today’s rule have only truthtelling in mind strikes me as far too sanguine to support acceptance of a rule that controls the hard reality of contested criminal trials. The majority expresses the common sense of the matter in saying that presentation of excluded evidence must sometimes be allowed because it “penalizes defendants for committing perjury.” Ante, at 314.

In some cases, of course, false testimony can result from faulty recollection. But the majority’s ironclad rule is one that applies regardless of the witness’ motives, and may be misused as a license to perjure. Even if the witness testifies in good faith, the defendant and his lawyer, who offer the testimony, know the facts. Indeed, it is difficult here to imagine the defense attorney’s reason for asking Henderson about petitioner’s hair color if he did not expect her to cast doubt on the eyewitness identification of petitioner by giving a description of petitioner’s hair color contrary to that contained in his own (suppressed) statement.

The suggestion that the threat of a perjury prosecution will provide sufficient deterrence to prevent false testimony, ante, *327at 314 (opinion of Brennan, J.); ante, at 320-321 (opinion of Stevens, J.), is not realistic. See generally Dunn v. United States, 442 U. S. 100, 108 (1979) (describing proof of perjury as “exceptionally difficult”). A heightened proof requirement applies in Illinois and other States, making perjury convictions difficult to sustain. See People v. Alkire, 321 Ill. 28, 151 N. E. 518 (1926); People v. Harrod, 140 Ill. App. 3d 96, 488 N. E. 2d 316 (1986). Where testimony presented on behalf of a friend or family member is involved, the threat that a future jury will convict the witness may be an idle one.

The damage to the truth-seeking process caused by the majority’s rule is certain to be great whether the testimony is perjured or merely false. In this case there can be little doubt of the falsity, since petitioner’s description of his own hair was at issue. And as a general matter the alternative to rebuttal is endorsement of judicial proceedings conducted in reliance on information known to be untrue. Suppressed evidence is likely to consist of either voluntary statements by the defendant himself or physical evidence. Both have a high degree of reliability, and testimony in direct conflict to such evidence most often will represent an attempt to place falsehoods before the jury.2

*328The suggestion that all this is so far beyond the control of the defendant that he will put on no defense is not supported. As to sympathetic witnesses, such as the family friend here, it should not be too hard to assure the witness does not volunteer testimony in contradiction of the facts. The defendant knows the content of the suppressed evidence. Even in cases- where the time for consultation is limited, the defense attorney can take care not to elicit contradicting testimony. And in the case of truly neutral witnesses, or witnesses hostile to the accused, it is hard to see the danger that they will present false testimony for the benefit of the defense.

The majority’s concerns may carry greater weight where contradicting testimony is elicited from a defense witness on cross-examination. In that situation there might be a concern that the prosecution would attempt to produce such testimony as the foundation to put excluded evidence before the jury. We have found that possibility insufficient to justify immunity for a defendant’s own false testimony on cross-examination. United States v. Havens, 446 U. S. 620 (1980). As to cross-examination of other witnesses, perhaps a different rule could be justified. Rather than wait for an appropriate case to consider this or similar measures, however, the majority opts for a wooden rule immunizing all defense testimony from rebuttal, without regard to knowledge that the testimony introduced at the behest of the defendant is false or perjured.

I also cannot agree that admission of excluded evidence on rebuttal would lead to the “disregard of . . . constitutional rights,” by law enforcement officers, ante, at 319, that the majority fears. This argument has been raised in our previous cases in this area of the law. See Havens, supra, at 633-634 (Brennan, J., dissenting); Hass, 420 U. S., at 725 (Brennan, J., dissenting); Harris, 401 U. S., at 232 (Brennan, J., dissenting). To date we have rejected it. Now the specter appears premised on an assumption that a single slip of the tongue by any defense witness will open the door to *329any suppressed evidence at the prosecutor’s disposal. If this were so, the majority’s concern that officers would be left with little to lose from conducting an illegal search would be understandable. And the argument might hold more force if, as the majority speculates, ante, at 319, police confront the temptation to seize evidence illegally “much if not most of the time” after gathering sufficient evidence to present proof of guilt beyond a reasonable doubt in the case in chief. Again, however, I disagree with the predictions.

It is unrealistic to say that the decision to make an illegal search turns on a precise calculation of the possibilities of rebuttal at some future trial. There is no reason to believe a police officer, unschooled in the law, will assess whether evidence already in his possession would suffice to survive a motion for acquittal following the case in chief. The officer may or may not even know the identity of the ultimate defendant.3 He certainly will not know anything about potential defense witnesses, much less what the content of their testimony might be. What he will know for certain is that evidence from an illegal search or arrest (which may well be crucial to securing a conviction) will be lost to the case in chief. Our earlier assessments of the marginal deterrent effect are applicable here. “Assuming that the exclusionary rule has a deterrent effect on proscribed police conduct, sufficient deterrence flows when the evidence in question is made un*330available to the prosecution in its case in chief.” Harris, supra, at 225.

In this case, the defense witness, one Jewel Henderson, testified that petitioner’s hair was black on the date of the offense. Her statement, perjured or not, should not have been offered to the jurors without giving them the opportunity to consider the unequivocal and contradicting description by the person whose own hair it was. I would allow the introduction of petitioner’s statement that his hair was red on the day of the shootings. The result is consistent with our line of cases from Walder to Havens and compelled by their reasoning.

The prosecution, it is true, did not limit itself to petitioner’s description of his hair color. It went beyond this to introduce petitioner’s statement that he went to the beauty shop to “change his appearance.” App. 11. The prosecutor used this statement to suggest that petitioner had a guilty mind and an intention to evade capture by disguise. This goes beyond what was necessary to rebut Henderson’s testimony and raises many of the concerns expressed in the majority opinion. Nonetheless, there was overwhelming evidence of petitioner’s guilt in this case, including the testimony of five eyewitnesses. In view of these circumstances, I agree with the Illinois Supreme Court that any error as to the additional statements or the prosecutor’s argument had no effect on petitioner’s trial and may be considered harmless.

Where the jury is misled by false testimony, otherwise subject to flat contradiction by evidence illegally seized, the protection of the exclusionary rule is “‘perverted into a license to use perjury by way of a defense, free from the risk of confrontation with prior inconsistent utterances.’” Havens, supra, at 626 (quoting Harris, supra, at 226). The perversion is the same where the perjury is by proxy. I would affirm the judgment of the Illinois Supreme Court.

Justice White

delivered the opinion of the Court.

In this case, we are called on to determine whether the interrogation of petitioner after his indictment violated his Sixth Amendment right to counsel.

I

Before dawn on August 21, 1983, petitioner and other members of the “Vice Lords” street gang became involved in a fight with members of a rival gang, the “Black Mobsters.” Some time after the. fight, a former member of the Black Mobsters, James Jackson, went to the home where the Vice Lords had fled. A second fight broke out there, with petitioner and three other Vice Lords beating Jackson severely. The Vice Lords then put Jackson into a car, drove to the end of a nearby street, and left him face down in a puddle of water. Later that morning, police discovered Jackson, dead, where he had been left.

That afternoon, local police officers obtained warrants for the arrest of the Vice Lords, on charges of battery and mob action, in connection with the first fight. One of the gang members who was arrested gave the police a statement concerning the first fight; the statement also implicated several of the Vice Lords (including petitioner) in Jackson’s murder. A few hours later, petitioner was apprehended. Petitioner was informed of his rights under Miranda v. Arizona, 384 U. S. 436 (1966), and volunteered to answer questions put to him by the police. Petitioner gave a statement concerning the initial fight between the rival gangs, but denied knowing anything *288about Jackson’s death. Petitioner was held in custody the following day, August 22, as law enforcement authorities completed their investigation of the Jackson murder.

On August 23, a Cook County grand jury indicted petitioner and two other gang members for the murder of James Jackson. Police Officer Michael Gresham, who had questioned petitioner earlier, removed him from the lockup where he was being held, and told petitioner that because he had been indicted he was being transferred to the Cook County jail. Petitioner asked Gresham which of the gang members had been charged with Jackson’s murder, and upon learning that one particular Vice Lord had been omitted from the indictments, asked: “[W]hy wasn’t he indicted, he did everything.” App. 7. Petitioner also began to explain that there was a witness who would support his account of the crime.

At this point, Gresham interrupted petitioner, and handed him a Miranda waiver form. The form contained five specific warnings, as suggested by this Court’s Miranda decision, to make petitioner aware of his right to counsel and of the consequences of any statement he might make to police.1 Gresham read the warnings aloud, as petitioner read along with him. Petitioner initialed each of the five warnings, and signed the waiver form. Petitioner then gave a lengthy statement to police officers concerning the Jackson murder; petitioner’s statement described in detail the role of each of the Vice Lords — including himself — in the murder of James Jackson.

Later that day, petitioner confessed involvement in the murder for a second time. This confession came in an inter*289view with Assistant State’s Attorney (ASA) George Smith. At the outset of the interview, Smith reviewed with petitioner the Miranda waiver he had previously signed, and petitioner confirmed that he had signed the waiver and understood his rights. Smith went through the waiver procedure once again: reading petitioner his rights, having petitioner initial each one, and sign a waiver form. In addition, Smith informed petitioner that he was a lawyer working with the police investigating the Jackson case. Petitioner then gave another inculpatory statement concerning the crime.

Before trial, petitioner moved to suppress his statements, arguing that they were obtained in a manner at odds with various constitutional guarantees. The trial court denied these motions, and the statements were used against petitioner at his trial. The jury found petitioner guilty of murder, and petitioner was sentenced to a 24-year prison term.

On appeal, petitioner argued that he had not “knowingly and intelligently” waived his Sixth Amendment right to counsel before he gave his uncounseled postindictment confessions. Petitioner contended that the warnings he received, while adequate for the purposes of protecting his Fifth Amendment rights as guaranteed by Miranda, did not adequately inform him of his Sixth Amendment right to counsel. The Illinois Supreme Court, however, rejected this theory, applying its previous decision in People v. Owens, 102 Ill. 2d 88, 464 N. E. 2d 261, cert. denied, 469 U. S. 963 (1984), which had held that Miranda warnings were sufficient to make a defendant aware of his Sixth Amendment right to counsel during postindictment questioning. People v. Thomas, 116 Ill. 2d 290, 298-300, 507 N. E. 2d 843, 846-847 (1987).

In reaching this conclusion, the Illinois Supreme Court noted that this Court had reserved decision on this question on several previous occasions2 and that the lower courts are *290divided on the issue. Id., at 299, 507 N. E. 2d, at 846. We granted this petition for certiorari, 484 U. S. 895 (1987), to resolve this split of authority and to address the issues we had previously left open.

II

There can be no doubt that petitioner had the right to have the assistance of counsel at his postindictment interviews with law enforcement authorities. Our cases make it plain that the Sixth Amendment guarantees this right to criminal defendants. Michigan v. Jackson, 475 U. S. 625, 629-630 (1986); Brewer v. Williams, 430 U. S. 387, 398-401 (1977); Massiah v. United States, 377 U. S. 201, 205-207 (1964).3 Petitioner asserts that the questioning that produced his incriminating statements violated his Sixth Amendment right to counsel in two ways.

A

Petitioner’s first claim is that because his Sixth Amendment right to counsel arose with his indictment, the police were thereafter barred from initiating a meeting with him. See Brief for Petitioner 30-31; Tr. of Oral Arg. 2, 9, 11, 17. He equates himself with a preindictment suspect who, while being interrogated, asserts his Fifth Amendment right to counsel; under Edwards v. Arizona, 451 U. S. 477 (1981), such a suspect may not be questioned again unless he initiates the meeting.

Petitioner, however, at no time sought to exercise his right to have counsel present. The fact that petitioner’s Sixth *291Amendment right came into existence with his indictment, i. e., that he had such a right at the time of his questioning, does not distinguish him from the preindictment interrogatee whose right to counsel is in existence and available for his exercise while he is questioned. Had petitioner indicated he wanted the assistance of counsel, the authorities’ interview with him would have stopped, and further questioning would have been forbidden (unless petitioner called for such a meeting). This was our holding in Michigan v. Jackson, supra, which applied Edwards to the Sixth Amendment context. We observe that the analysis in Jackson is rendered wholly unnecessary if petitioner’s position is correct: under petitioner’s theory, the officers in Jackson would have been completely barred from approaching the accused in that case unless he called for them. Our decision in Jackson, however, turned on the fact that the accused “ha[d] asked for the help of a lawyer” in dealing with the police. Jackson, supra, at 631, 633-635.

At bottom, petitioner’s theory cannot be squared with our rationale in Edwards, the case he relies on for support. Edwards rested on the view that once “an accused . . . ha[s] expressed his desire to deal with the police only through counsel” he should “not [be] subject to further interrogation by the authorities until counsel has been made available to him, unless the accused himself initiates further communication.” Edwards, supra, at 484-485; cf. also Michigan v. Mosley, 423 U. S. 96, 104, n. 10 (1975). Preserving the integrity of an accused’s choice to communicate with police only through counsel is the essence of Edwards and its progeny— not barring an accused from making an initial election as to whether he will face the State’s officers during questioning with the aid of counsel, or go it alone. If an accused “knowingly and intelligently” pursues the latter course, we see no reason why the uncounseled statements he then makes must be excluded at his trial.

*292B

Petitioner’s principal and more substantial claim is that questioning him without counsel present violated the Sixth Amendment because he did not validly waive his right to have counsel present during the interviews. Since it is clear that after the Miranda warnings were given to petitioner, he not only voluntarily answered questions without claiming his right to silence or his right to have a lawyer present to advise him but also executed a written waiver of his right to counsel during questioning, the specific issue posed here is whether this waiver was a “knowing and intelligent” waiver of his Sixth Amendment right.4 See Brewer v. Williams, supra, at 401, 404; Johnson v. Zerbst, 304 U. S. 458, 464-465 (1938).

In the past, this Court has held that a waiver of the Sixth Amendment right to. counsel is valid only when it reflects “an intentional relinquishment or abandonment of a known right or privilege.” Johnson v. Zerbst, supra, at 464. In other words, the accused must “kno[w] what he is doing” so that “his choice is made with eyes open.” Adams v. United States ex rel. McCann, 317 U. S. 269, 279 (1942). In a case arising under the Fifth Amendment, we described this requirement as “a full awareness of both the nature of the right being abandoned and the consequences of the decision to abandon it.” Moran v. Burbine, 475 U. S. 412, 421 (1986). Whichever of these formulations is used, the key inquiry in a case such as this one must be: Was the accused, who waived his Sixth Amendment rights during postindictment questioning, made sufficiently aware of his right to have counsel present during the questioning, and of the possible conse*293quences of a decision to forgo the aid of counsel? In this case, we are convinced that by admonishing petitioner with the Miranda warnings, respondent has met this burden and that petitioner’s waiver of his right to counsel at the questioning was valid.5

First, the Miranda warnings given petitioner made him aware of his right to have counsel present during the questioning. By telling petitioner that he had a right to consult with an attorney, to have a lawyer present while he was questioned, and even to have a lawyer appointed for him if he could not afford to retain one on his own, Officer Gresham and ASA Smith conveyed to petitioner the sum and substance of the rights that the Sixth Amendment provided him. “Indeed, it seems self-evident that one who is told he” has such rights to counsel “is in a curious posture to later complain” that his waiver of these rights was unknowing. Cf. United States v. Washington, 431 U. S. 181, 188 (1977). There is little more petitioner could have possibly been told in an effort to satisfy this portion of the waiver inquiry.

Second, the Miranda warnings also served to make petitioner aware of the consequences of a decision by him to waive his Sixth Amendment rights during postindictment questioning. Petitioner knew that any statement that he made could be used against him in subsequent criminal proceedings. This is the ultimate adverse consequence petitioner could have suffered by virtue of his choice to make *294uncounseled admissions to the authorities. This warning also sufficed — contrary to petitioner’s claim here, see Tr. of Oral Arg. 7-8 — to let petitioner know what a lawyer could “do for him” during the postindictment questioning: namely, advise petitioner to refrain from making any such statements.6 By knowing what could be done with any statements he might make, and therefore, what benefit could be obtained by having the aid of counsel while making such statements, petitioner was essentially informed of the possible consequences of going without counsel during questioning. If petitioner nonetheless lacked “a full and complete appreciation of all of the consequences flowing” from his waiver, it does not defeat the State’s showing that the information it provided to him satisfied the constitutional minimum. Cf. Oregon v. Elstad, 470 U. S. 298, 316-317 (1985).

Our conclusion is supported by petitioner’s inability, in the proceedings before this Court, to articulate with precision what additional information should have been provided to him before he would have been competent to waive his right to counsel. All that petitioner’s brief and reply brief suggest is petitioner-should have been made aware of his “right under the Sixth Amendment to the broad protection of counsel” — a rather nebulous suggestion — and the “gravity of [his] situation.” Reply Brief for Petitioner 13; see Brief for Petitioner 30-31. But surely this latter “requirement” (if it is one) was met when Officer Gresham informed petitioner that he had been formally charged with the murder of James Jackson. *295See n. 8, infra. Under close questioning on this same point at argument, petitioner likewise failed to suggest any meaningful additional information that he should have been, but was not, provided in advance of his decision to waive his right to counsel.7 The discussions found in favorable court decisions, on which petitioner relies, are similarly lacking.8

*296As a general matter, then, an accused who is admonished with the warnings prescribed by this Court in Miranda, 384 U. S., at 479, has been sufficiently apprised of the nature of his Sixth Amendment rights, and of the consequences of abandoning those rights, so that his waiver on this basis will be considered a knowing and intelligent one.9 We feel that *297our conclusion in a recent Fifth Amendment case is equally apposite here: “Once it is determined that a suspect's decision not to rely on his rights was uncoerced, that he at all times knew he could stand mute and request a lawyer, and that he was aware of the State’s intention to use his statements to secure a conviction, the analysis is complete and the waiver is valid as a matter of law.” See Moran v. Burbine, 475 U. S., at 422-423.

C

We consequently reject petitioner’s argument, which has some acceptance from courts and commentators,10 that since “the sixth amendment right [to counsel] is far superior to that of the fifth amendment right” and since “[t]he greater the right the greater the loss from a waiver of that right,” waiver of an accused’s Sixth Amendment right to counsel should be “more difficult” to effectuate than waiver of a suspect’s Fifth Amendment rights. Brief for Petitioner 23. While our cases have recognized a “difference” between the Fifth Amendment and Sixth Amendment rights to counsel, and the “policies” behind these constitutional guarantees,11 we have never suggested that one right is “superior” or “greater” than the other, nor is there any support in our cases for the notion that be*298cause a Sixth Amendment right may be involved, it is more difficult to waive than the Fifth Amendment counterpart.

Instead, we have taken a more pragmatic approach to the waiver question — asking what purposes a lawyer can serve at the particular stage of the proceedings in question, and what assistance he could provide to an accused at that stage — to determine the scope of the Sixth Amendment right to counsel, and the type of warnings and procedures that should be required before a waiver of that right will be recognized.

At one end of the spectrum, we have concluded there is no Sixth Amendment right to counsel whatsoever at a postin-dictment photographic display identification, because this procedure is not one at which the accused “require[s] aid in coping with legal problems or assistance in meeting his adversary.” See United States v. Ash, 413 U. S. 300, 313-320 (1973). At the other extreme, recognizing the enormous importance and role that an attorney plays at a criminal trial, we have imposed the most rigorous restrictions on the information that must be conveyed to a defendant, and the procedures that must be observed, before permitting him to waive his right to counsel at trial. See Faretta v. California, 422 U. S. 806, 835-836 (1975); cf. Von Moltke v. Gillies, 332 U. S. 708, 723-724 (1948). In these extreme cases, and in others that fall between these two poles, we have defined the scope of the right to counsel by a pragmatic assessment of the usefulness of counsel to the accused at the particular proceeding, and the dangers to the accused of proceeding without counsel. An accused’s waiver of his right to counsel is “knowing” when he is made aware of these basic facts.

Applying this approach, it is our view that whatever warnings suffice for Miranda’s purposes will also be sufficient in the context of postindictment questioning. The State’s decision to take an additional step and commence formal adversarial proceedings against the accused does not substantially increase the value of counsel to the accused at questioning, or expand the limited purpose that an attorney serves when the *299accused is questioned by authorities. With respect to this inquiry, we do not discern a substantial difference between the usefulness of a lawyer to a suspect during custodial interrogation, and his value to an accused at postindictment questioning.12

Thus, we require a more searching or formal inquiry before permitting an accused to waive his right to counsel at trial than we require for a Sixth Amendment waiver during post-indictment questioning — not because postindictment questioning is “less important” than a trial (the analysis that petitioner’s “hierarchical” approach would suggest) — but because the full “dangers and disadvantages of self-representation,” Faretta, supra, at 835, during questioning are less substantial and more obvious to an accused than they are at trial.13 Because the role of counsel at questioning is relatively simple and limited, we see no problem in having a waiver procedure at that stage which is likewise simple and limited. So long as the accused is made aware of the “dangers and disadvantages *300of self-representation” during postindictment questioning, by use of the Miranda warnings, his waiver of his Sixth Amendment right to counsel at such questioning is “knowing and intelligent.”

Ill

Before confessing to the murder of James Jackson, petitioner was meticulously informed by authorities of his right to counsel, and of the consequences of any choice not to exercise that right. On two separate occasions, petitioner elected to forgo the assistance of counsel, and speak directly to officials concerning his role in the murder. Because we believe that petitioner’s waiver of his Sixth Amendment rights was “knowing and intelligent,” we find no error in the decision of the trial court to permit petitioner’s confessions to be used against him. Consequently, the judgment of the Illinois Supreme Court is

Affirmed.

Justice Blackmun,

dissenting.

I agree with most of what Justice Stevens says in his dissenting opinion, post, p. 301. I, however, merely would hold that after formal adversary proceedings against a defendant have been commenced, the Sixth Amendment mandates that the defendant not be “‘subject to further interrogation by the authorities until counsel has been made available to him, unless the accused himself initiates further communication, exchanges, or conversations with the police.’” Michigan v. Jackson, 475 U. S. 625, 626 (1986), quoting Edwards v. Arizona, 451 U. S. 477, 484-485 (1981).

The Court’s majority concludes, ante, at 290-291: “The fact that petitioner’s Sixth Amendment right came into existence with his indictment . . . does not distinguish him from the preindictment interrogatee whose right to counsel is in existence and available for his exercise while he is questioned.” I must disagree. “[W]hen the Constitution grants protection against criminal proceedings without the assistance of coun*301sel, counsel must be furnished whether or not the accused requested the appointment of counsel.” Carnley v. Cochran, 369 U. S. 506, 513 (1962) (internal quotations omitted). In my view, the Sixth Amendment does not allow the prosecution to take undue advantage of any gap between the commencement of the adversary process and the time at which counsel is appointed for a defendant.

Justice Stevens,

with whom Justice Brennan and Justice Marshall join, dissenting.

The Court should not condone unethical forms of trial preparation by prosecutors or their investigators. In civil litigation it is improper for a lawyer to communicate with his or her adversary’s client without either notice to opposing counsel or the permission of the court.1 An attempt to obtain evidence for use at trial by going behind the back of one’s adversary would be not only a serious breach of professional ethics but also a manifestly unfair form of trial practice. In the criminal context, the same ethical rules apply and, in my opinion, notions of fairness that are at least as demanding should also be enforced.

After a jury has been empaneled and a criminal trial is in progress, it would obviously be improper for the prosecutor to conduct a private interview with the defendant for the pur*302pose of obtaining evidence to be used against him at trial. By “private interview” I mean, of course, an interview initiated by the prosecutor, or his or her agents, without notice to the defendant’s lawyer and without the permission of the court. Even if such an interview were to be commenced by giving the defendant the five items of legal advice that are mandated by Miranda, see ante, at 288, n. 1,1 have no doubt that this Court would promptly and unanimously condemn such a shabby practice. As our holding in Michigan v. Jackson, 475 U. S. 625 (1986), suggests, such a practice would not simply constitute a serious ethical violation, but would rise to the level of an impairment of the Sixth Amendment right to counsel.2

*303The question that this case raises, therefore, is at what point in the adversary process does it become impermissible for the prosecutor, or his or her agents, to conduct such private interviews with the opposing party? Several alternatives are conceivable: when the trial commences, when the defendant has actually met and accepted representation by his or her appointed counsel, when counsel is appointed, or when the adversary process commences. In my opinion, the Sixth Amendment right to counsel demands that a firm and unequivocal line be drawn at the point at which adversary proceedings commence.

In prior cases this Court has used strong language to emphasize the significance of the formal commencement of adversary proceedings. Such language has been employed to explain decisions denying the defendant the benefit of the protection of the Sixth Amendment in preindictment settings, but an evenhanded interpretation of the Amendment would support the view that additional protection should automatically attach the moment the formal proceed*304ings begin. One such example is Kirby v. Illinois, 406 U. S. 682 (1972), in which the Court concluded that the general rule requiring the presence of counsel at pretrial, lineup identifications, see United States v. Wade, 388 U. S. 218 (1967); Gilbert v. California, 388 U. S. 263 (1967), should not extend to protect custodial defendants not yet formally charged. Justice Stewart’s plurality opinion explained the significance of the formal charge:

“The initiation of judicial criminal proceedings is far from a mere formalism. It is the starting point of our whole system of adversary criminal justice. For it is only then that the government has committed itself to prosecute, and only then that the adverse positions of government and defendant have solidified. It is then that a defendant finds himself faced with the prosecuto-rial forces of organized society, and immersed in the intricacies of substantive and procedural criminal law. It is this point, therefore, that marks the commencement of the ‘criminal prosecutions’ to which alone the explicit guarantees of the Sixth Amendment are applicable. See Powell v. Alabama, 287 U. S., at 66-71; Massiah v. United States, 377 U. S. 201; Spano v. New York, 360 U. S. 315, 324 (Douglas, J., concurring).” 406 U. S., at 689-690 (footnote omitted).

Similarly, in United States v. Gouveia, 467 U. S. 180 (1984), we relied upon the significance of the absence of a formal charge in concluding that the Sixth Amendment does not require the appointment of counsel for indigent prison inmates confined in administrative detention while authorities investigate their possible involvement in criminal activity. Again the Court noted that “given the plain language of the Amendment and its purpose of protecting the unaided layman at critical confrontations with his adversary, our conclusion that the right to counsel attaches at the initiation of adversary ju*305dicial criminal proceedings ‘is far from a mere formalism.’ Kirby v. Illinois, 406 U. S., at 689.” Id., at 189.

Most recently, in Moran v. Burbine, 475 U. S. 412 (1986), the Court upheld a waiver of the right to counsel in a pretrial context even though the waiver “would not be valid” if the same situation had arisen after indictment, see ante, at 296-297, n. 9. In the Moran opinion, the Court explained:

“It is clear, of course, that, absent a valid waiver, the defendant has the right to the presence of an attorney during any interrogation occurring after the first formal charging proceeding, the point at which the Sixth Amendment right to counsel initially attaches. United States v. Gouveia, 467 U. S. 180, 187 (1984); Kirby v. Illinois, 406 U. S. 682, 689 (1972) (opinion of Stewart, J.). See Brewer v. Williams, 430 U. S., at 400-401. And we readily agree that once the right has attached, it follows that the police may not interfere with the efforts of a defendant’s attorney to act as a ‘ “medium” between [the suspect] and the State’ during the interrogation. Maine v. Moulton, 474 U. S. 159, 176 (1985); see Brewer v. Williams, supra, at 401, n. 8. The difficulty for respondent is that the interrogation sessions that yielded the inculpatory statements took place before the initiation of ‘adversary judicial proceedings.’ United States v. Gouveia, supra, at 192.” 475 U. S., at 428.

Today, however, in reaching a decision similarly favorable to the interest in law enforcement unfettered by process concerns, the Court backs away from the significance previously attributed to the initiation of formal proceedings. In the majority’s view, the purported waiver of counsel in this case is properly equated with that of an unindicted suspect. Yet, as recognized in Kirby, Gouveia, and Moran, important differ-*306enees separate the two.3 The return of an indictment, or like instrument, substantially alters the relationship between the state and the accused. Only after a formal accusation has “the government. . . committed itself to prosecute, and only then [have] the adverse positions of government and defendant . . . solidified.” Kirby, 406 U. S., at 689. Moreover, the return of an indictment also presumably signals the government’s conclusion that it has sufficient evidence to establish a prima facie case. As a result, any further interrogation can only be designed to buttress the government’s case; authorities are no longer simply attempting “ ‘to solve a crime.’” United States v. Mohabir, 624 F. 2d 1140, 1148 (CA2 1980) (quoting People v. Waterman, 9 N. Y. 2d 561, 565, 175 N. E. 2d 445, 447 (1961)); see also Moran v. Burbine, 475 U. S., at 430. Given the significance of the initiation of formal proceedings and the concomitant shift in the relationship between the state and the accused, I think it quite wrong to suggest that Miranda warnings — of for that *307matter, any warnings offered by an adverse party — provide a sufficient basis for permitting the undoubtedly prejudicial— and, in my view, unfair — practice of permitting trained law enforcement personnel and prosecuting attorneys to communicate with as-of-yet unrepresented criminal defendants.

It is well settled that there is a strong presumption against waiver of Sixth Amendment protections, see Michigan v. Jackson, 475 U. S., at 633; Von Moltke v. Gillies, 332 U. S. 708, 723 (1948) (plurality opinion); Johnson v. Zerbst, 304 U. S. 458, 464 (1938), and that a waiver may only be accepted if made with full awareness of “the dangers and disadvantages of self-representation,” Faretta v. California, 422 U. S. 806, 835 (1975); see also Adams v. United States ex rel. McCann, 317 U. S. 269, 279 (1942) (accused “may waive his Constitutional right to assistance of counsel if he knows what he is doing and his choice is made with eyes open”). Warnings offered by an opposing party, whether detailed or cursory, simply cannot satisfy this high standard.

The majority premises its conclusion that Miranda warnings lay a sufficient basis for accepting a waiver of the right to counsel on the. assumption that those warnings make clear to an accused “what a lawyer could ‘do for him’ during the postindictment questioning: namely, advise [him] to refrain from making any [incriminating] statements.” Ante, at 294 (footnote omitted).4 Yet, this is surely a gross understatement of the disadvantage of proceeding without a lawyer and *308an understatement of what a defendant must understand to make a knowing waiver.5 The Miranda warnings do not, for example, inform the accused that a lawyer might examine the indictment for legal sufficiency before submitting his or her client to interrogation or that a lawyer is likely to be considerably more skillful at negotiating a plea bargain and that such negotiations may be most fruitful if initiated prior to any interrogation. Rather, the warnings do not even go so far as to explain to the accused the nature of the charges pending against him — advice that a court would insist upon before allowing a defendant to enter a guilty plea with or' without the presence of an attorney, see Henderson v. Morgan, 426 U. S. 637 (1976). Without defining precisely the nature of the inquiry required to establish a valid waiver of the Sixth Amendment right to counsel, it must be conceded that at least minimal advice is necessary — the accused must be told of the “dangers and disadvantages of self-representation.”

Yet, once it is conceded that certain advice is required and that after indictment the adversary relationship between the state and the accused has solidified, it inescapably follows *309that a prosecutor may not conduct private interviews with a charged defendant. As at least one Court of Appeals has recognized, there are ethical constraints that prevent a prosecutor from giving legal advice to an uncounseled adversary.6 Thus, neither the prosecutor nor his or her agents can ethically provide the unrepresented defendant with the kind of advice that should precede an evidence-gathering interview after formal proceedings have been commenced. Indeed, in my opinion even the Miranda warnings themselves are a species of legal advice that is improper when given by the prosecutor after indictment.

Moreover, there are good reasons why such advice is deemed unethical, reasons that extend to the custodial, post-indictment setting with unequaled strength. First, the offering of legal advice may lead an accused to underestimate the prosecuting authorities’ true adversary posture. For an incarcerated defendant — in this case, a 17-year-old who had been in custody for 44 hours at the time he was told of the *310indictment — the assistance of someone to explain why he is being held, the nature of the charges against him, and the extent of his legal rights, may be of such importance as to overcome what is perhaps obvious to most, that the prosecutor is a foe and not a friend. Second, the adversary posture of the parties, which is not fully solidified until formal charges are brought, will inevitably tend to color the advice offered. As hard as a prosecutor might try, I doubt that it is possible for one to wear the hat of an effective adviser to a criminal defendant while at the same time wearing the hat of a law enforcement authority. Finally, regardless of whether or not the accused actually understands the legal and factual issues involved and the state’s role as an adversary party, advice offered by a lawyer (or his or her agents) with such an evident conflict of interest cannot help but create a public perception of unfairness and unethical conduct. And as we held earlier this Term, “courts have an independent interest in ensuring that criminal trials are conducted within the ethical standards of the profession and that legal proceedings appear fair to all who observe them.” Wheat v. United States, 486 U. S. 153, 160 (1988). This interest is a factor that may be considered in deciding whether to override a defendant’s waiver of his or her Sixth Amendment right to conflict-free representation, see ibid., and likewise, should be considered in determining whether a waiver based on advice offered by the criminal defendant’s adversary is ever appropriate.7

In sum, without a careful discussion of the pitfalls of proceeding without counsel, the Sixth Amendment right cannot properly be waived. An adversary party, moreover, cannot adequately provide such advice. As a result, once the right to counsel attaches and the adversary relationship between *311the state and the accused solidifies, a prosecutor cannot conduct a private interview with an accused party without “dilut[ing] the protection afforded by the right to counsel,” Maine v. Moulton, 474 U. S. 159, 171 (1985). Although this ground alone is reason enough to never permit such private interviews, the rule also presents the added virtue of drawing a clear and easily identifiable line at the point between the investigatory and adversary stages of a criminal proceeding. Such clarity in definition of constitutional rules that govern criminal proceedings is important to the law enforcement profession as well as to the private citizen. See Arizona v. Roberson, 486 U. S. 675 (1988). It is true, of course, that the interest in effective law enforcement would benefit from an opportunity to engage in incommunicado questioning of defendants who, for reasons beyond their control, have not been able to receive the legal advice from counsel to which they are constitutionally entitled. But the Court’s singleminded concentration on that interest might also lead to the toleration of similar practices at any stage of the trial. I think it clear that such private communications are intolerable not simply during trial, but at any point after adversary proceedings have commenced.

I therefore respectfully dissent.

*173Justice Scalia

delivered the opinion of the Court.

This case presents the question whether an accused’s invocation of his Sixth Amendment right to counsel during a judicial proceeding constitutes an invocation of his Miranda right to counsel.

I

Petitioner Paul McNeil was arrested in Omaha, Nebraska, in May 1987, pursuant to a warrant charging him with an armed robbery in West Allis, Wisconsin, a suburb of Milwaukee. Shortly after his arrest, two Milwaukee County deputy sheriffs arrived in Omaha to retrieve him. After advising him of his Miranda rights, the deputies sought to question him. He refused to answer any questions, but did not request an attorney. The deputies promptly ended the interview.

Once back in Wisconsin, petitioner was brought before a Milwaukee County Court Commissioner on the armed robbery charge. The Commissioner set bail and scheduled a preliminary examination. An attorney from the Wisconsin Public Defender’s Office represented petitioner at this initial appearance.

Later that evening, Detective Joseph Butts of the Milwaukee County Sheriff’s Department visited petitioner in jail. Butts had been assisting the Racine County, Wisconsin, police in their investigation of a murder, attempted murder, and armed burglary in the town of Caledonia; petitioner was a suspect. Butts advised petitioner of his Miranda rights, and petitioner signed a form waiving them. In this *174first interview, petitioner did not deny knowledge of the Caledonia crimes, but said that he had not been involved.

Butts returned two donia. He again began the encounter by advising petitioner of his Miranda rights and providing a waiver form. Petitioner placed his initials next to each of the warnings and signed the form. This time, petitioner admitted that he had been involved in the Caledonia crimes, which he described in detail. He also implicated two other men, Willie Pope and Lloyd Crowley. The statement was typed up by a detective and given to petitioner to review. Petitioner placed his initials next to every reference to himself and signed every page.

Butts and the Caledonia having in the meantime found and questioned Pope, who convinced them that he had not been involved in the Caledonia crimes. They again began the interview by administering the Miranda warnings and obtaining petitioner’s signature and initials on the waiver form. Petitioner acknowledged that he had lied about Pope’s involvement to minimize his own role in the Caledonia crimes and provided another statement recounting the events, which was transcribed, signed, and initialed as before.

The following day, petitioner was the Caledonia crimes and transferred to that jurisdiction. His pretrial motion to suppress the three incriminating statements was denied. He was convicted of second-degree murder, attempted first-degree murder, and armed robbery, and sentenced to 60 years in prison.

On appeal, petitioner argued that the trial court’s refusal to suppress the statements was reversible error. He contended that his courtroom appearance with an attorney for the West Allis crime constituted an invocation of the Miranda right to counsel, and that any subsequent waiver of that right during police-initiated questioning regarding any offense was invalid. Observing that the State’s Supreme *175Court had never addressed this issue, the Court of Appeals certified to that court the following question:

“Does an accused’s request for counsel at an initial appearance on a charged offense constitute an invocation of his fifth amendment right to counsel that precludes police-initiated interrogation on unrelated, uncharged offenses?” App. 16.

The Wisconsin Supreme Court answered “no.” 155 Wis. 2d 24, 454 N. W. 2d 742 (1990). We granted certiorari, 498 U. S. 937 (1990).

II

The Sixth Amendment provides that “[i]n all criminal prosecutions, the accused shall enjoy the right ... to have the Assistance of Counsel for his defence.” In Michigan v. Jackson, 475 U. S. 625 (1986), we held that once this right to counsel has attached and has been invoked, any subsequent waiver during a police-initiated .custodial interview is ineffective. It is undisputed, and we accept for purposes of the present case, that at the time petitioner provided the incriminating statements at issue, his Sixth Amendment right had attached and had been invoked with respect to the West Allis armed robbery, for which he had been formally charged.

Sixth Amendment right, however, is offense specific. It cannot be invoked once for all future prosecutions, for it does not attach until a prosecution is commenced, that is, “ ‘at or after the initiation of adversary judicial criminal proceedings — whether by way of formal charge, preliminary hearing, indictment, information, or arraignment.’” United States v. Gouveia, 467 U. S. 180, 188 (1984) (quoting Kirby v. Illinois, 406 U. S. 682, 689 (1972) (plurality opinion)). And just as the right is offense specific, so also its Michigan v. Jackson effect of invalidating subsequent waivers in police-initiated interviews is offense specific.

“The police have an interest... in investigating new or additional crimes [after an individual is formally charged *176with one crime.] . . . [T]o exclude evidence pertaining to charges as to which the Sixth Amendment right to counsel had not attached at the time the evidence was obtained, simply because other charges were pending at that time, would unnecessarily frustrate the public’s interest in the investigation of criminal activities. . . .” Maine v. Moulton, 474 U. S. 159, 179-180 (1985).

“Incriminating statements pertaining to other crimes, as to which the Sixth Amendment right has not yet attached, are, of course, admissible at a trial of those offenses.” Id., at 180, n. 16.

See also Moran v. Burbine, 475 U. S. 412, 431 (1986). Because petitioner provided the statements at issue here before his Sixth Amendment right to counsel with respect to the Caledonia offenses had been (or even could have been) invoked, that right poses no bar to the admission of the statements in this case.

Petitioner relies, however, upon a different “right to counsel,” found not in the text of the Sixth Amendment, but in this Court’s jurisprudence relating to the Fifth Amendment guarantee that “[n]o person . . . shall be compelled in any criminal case to be a witness against himself.” In Miranda v. Arizona, 384 U. S. 436 (1966), we established a number of prophylactic rights designed to counteract the “inherently compelling pressures” of custodial interrogation, including the right to have counsel present. Miranda did not hold, however, that those rights could not be waived. On the contrary, the opinion recognized that statements elicited during custodial interrogation would be admissible if the prosecution could establish that the suspect “knowingly and intelligently waived his privilege against self-incrimination and his right to retained or appointed counsel.” Id., at 475.

In Edwards v. Arizona, 451 U. S. 477 (1981), we established a second layer of prophylaxis for the Miranda right to counsel: Once a suspect asserts the right, not only must the *177current interrogation cease, but he may not be approached for further interrogation “until counsel has been made available to him,” 451 U. S., at 484-485—which means, we have most recently held, that counsel must be present, Minnick v. Mississippi, 498 U. S. 146 (1990). If the police do subsequently initiate an encounter in the absence of counsel (assuming there has been no break in custody), the suspect’s statements are presumed involuntary and therefore inadmissible as substantive evidence at trial, even where the suspect executes a waiver and his statements would be considered voluntary under traditional standards. This is “designed to prevent police from badgering a defendant into waiving his previously asserted Miranda rights,” Michigan v. Harvey, 494 U. S. 344, 350 (1990). The Edwards rule, moreover, is not offense specific: Once a suspect invokes the Miranda right to counsel for interrogation regarding one offense, he may not be reapproached regarding any offense unless counsel is present. Arizona v. Roberson, 486 U. S. 675 (1988).

Having described the nature and effects of both the Sixth Amendment right to counsel and the Miranda-Edwards “Fifth Amendment” right to counsel, we come at last to the issue here: Petitioner seeks to prevail by combining the two of them. He contends that, although he expressly waived his Miranda right to counsel on every occasion he was interrogated, those waivers were the invalid product of impermissible approaches, because his prior invocation of the offense-specific Sixth Amendment right with regard to the West Allis burglary was also an invocation of the nonoffense-specific Miranda-Edwards right. We think that is false as a matter of fact and inadvisable (if even permissible) as a contrary-to-fact presumption of policy.

As to the former: The purpose of the Sixth Amendment counsel guarantee — and hence the purpose of invoking it — is to “protec[t] the unaided layman at critical confrontations” with his “expert adversary,” the government, after “the ad*178verse positions of government and defendant have solidified” with respect to a particular alleged crime. Gouveia, 467 U. S., at 189. The purpose of the Miranda-Edwards guarantee, on the other hand — and hence the purpose of invoking it — is to protect a quite different interest: the suspect’s “desire to deal with the police only through counsel,” Edwards, supra, at 484. This is in one respect narrower than the interest protected by the Sixth Amendment guarantee (because it relates only to custodial interrogation) and in another respect broader (because it relates to interrogation regarding any suspected crime and attaches whether or not the “adversarial relationship” produced by a pending prosecution has yet arisen). To invoke the Sixth Amendment interest is, as a matter of fact, not to invoke the Miranda-Edwards interest. One might be quite willing to speak to the police without counsel present concerning many matters, but not the matter under prosecution. It can be said, perhaps, that it is likely that one who has asked for counsel’s assistance in defending against a prosecution would want counsel present for all custodial interrogation, even interrogation unrelated to the charge. That is not necessarily true, since suspects often believe that they can avoid the laying of charges by demonstrating an assurance of innocence through frank and unassisted answers to questions. But even if it were true, the likelihood that a suspect would wish counsel to be present is not the test for applicability of Edwards. The rule of that case applies only when the suspect “ha[s] expressed” his wish for the particular sort of lawyerly assistance that is the subject of Miranda. Edwards, supra, at 484 (emphasis added). It requires, at a minimum, some statement that can reasonably be construed to be an expression of a desire for the assistance of an attorney in dealing with custodial interrogation by the police. Requesting the assistance of an attorney at a bail hearing does not bear that construction. “[T]o find that [the defendant] invoked his Fifth Amendment right to counsel on the present charges merely by requesting *179the appointment of counsel at his arraignment on the unrelated charge is to disregard the ordinary meaning of that request.” State v. Stewart, 113 Wash. 2d 462, 471, 780 P. 2d 844, 849 (1989), cert. denied, 494 U. S. 1020 (1990).

Our holding in Michigan v. Jackson, 475 U. S. 625 (1986), does not, as petitioner asserts, contradict the foregoing distinction; to the contrary, it rests upon it. That case, it will be recalled, held that after the Sixth Amendment right to counsel attaches and is invoked, any statements obtained from the accused during subsequent police-initiated custodial questioning regarding the charge at issue (even if the accused purports to waive his rights) are inadmissible. The State in Jackson opposed that outcome on the ground that assertion of the Sixth Amendment right to counsel did not realistically constitute the expression (as Edivards required) of a wish to have counsel present during custodial interrogation. See 475 U. S., at 632-633. Our response to that contention was not that it did constitute such an expression, but that it did not have to, since the relevant question was not whether the Miranda “Fifth Amendment” right had been asserted, but whether the Sixth Amendment right to counsel had been waived. We said that since our “settled approach to questions of waiver requires us to give a broad, rather than a narrow, interpretation to a defendant’s request for counsel, ... we presume that the defendant requests the lawyer’s services at every critical stage of the prosecution.” 475 U. S., at 633 (emphasis added). The holding of Jackson implicitly rejects any equivalence in fact between invocation of the Sixth Amendment right to counsel and the expression necessary to trigger Edivards. If such invocation constituted a real (as opposed to merely a legally presumed) request for the assistance of counsel in custodial interrogation, it would have been quite unnecessary for Jackson to go on to establish, as it did, a new Sixth Amendment rule of no police-*180initiated interrogation; we could simply have cited and relied upon Edwards. 1

There remains to though the assertion of the Sixth Amendment right to counsel does not in fact imply an assertion of the Miranda “Fifth Amendment” right, we should declare it to be. such as a matter of sound policy. Assuming we have such an expansive power under the Constitution, it would not wisely be exercised. Petitioner’s proposed rule has only insignificant advantages. If a suspect does not wish to communicate with the police except through an attorney, he can simply tell them that when they give him the Miranda warnings. There is not the remotest chance that he will feel “badgered” by their asking to talk to him without counsel present, since the subject will not be the charge on which he has already requested counsel’s assistance (for in that event Jackson would preclude initiation of the interview) and he will not have rejected uncounseled interrogation on any subject before (for in that event Edwards would preclude initiation of the interview). The proposed rule would, however, seriously impede effective law enforcement. The Sixth Amendment right to *181counsel attaches at the first formal proceeding against an accused, and in most States, at least with respect to serious offenses, free counsel is made available at that time and ordinarily requested. Thus, if we were to adopt petitioner’s rule, most persons in pretrial custody for serious offenses would be unapproachable by police officers suspecting them of involvement in other crimes, even though they have never expressed any unwillingness to be questioned. Since the ready ability to obtain uncoerced confessions is not an evil but an unmitigated good, society would be the loser. Admissions of guilt resulting from valid Miranda waivers “are more than merely ‘desirable’; they are essential to society’s compelling interest in finding, convicting, and punishing those who violate the law.” Moran, 475 U. S., at 426 (citation omitted).2

Petitioner urges upon us the desirability of providing a “clear and unequivocal” guideline for the police: no police-initiated questioning of any person in custody who has requested counsel to assist him in defense or in interrogation. But the police do not need our assistance to establish such a *182guideline; they are free, if they wish, to adopt it on their own. Of course it is our task to establish guidelines for judicial review. We like them to be “clear and unequivocal,” see, e. g., Roberson, 486 U. S., at 681-682, but only when they guide sensibly and in a direction we are authorized to go. Petitioner’s proposal would in our view do much more harm than good, and is not contained within, or even in furtherance of, the Sixth Amendment’s right to counsel or the Fifth Amendment’s right against compelled self-incrimination.3

* * *

"This Court is forever adding new stories to the temples of constitutional law, and the temples have a way of collapsing when one story too many is added.” Douglas v. Jeannette, 319 U. S. 157, 181 (1943) (opinion of Jackson, J.). We decline to add yet another story to Miranda. The judgment of the Wisconsin Supreme Court is

Affirmed.

*183Justice Kennedy,

concurring.

I join the opinion of the Court in all respects. Its sensible recognition that invocation of the Sixth Amendment right to counsel is specific to the offense in question should apply as well to requests for counsel under the Fifth Amendment. See Arizona v. Roberson, 486 U. S. 675, 688 (1988) (Kennedy, J., dissenting). For those in custody, Edwards v. Arizona, 451 U. S. 477 (1981), and its progeny go far to protect an individual who desires the assistance of counsel during interrogation. Limiting the extraordinary protections of Edwards to a particular investigation would not increase the risk of confessions induced by official efforts to wear down the will of a suspect. Having adopted an offense-specific rule for invocation of the Sixth Amendment right to counsel, the Court should devote some attention to bringing its Fifth and Sixth Amendment jurisprudence into a logical alignment, and should give uniform, fair, and workable guidelines for the criminal justice system.

Even if petitioner had invoked his Fifth Amendment right with respect to the West Allis armed robbery, I do not believe the authorities should have been prohibited from ques-' tioning him in connection with the Caledonia offenses.

Justice Stevens,

with whom Justice Marshall and Justice Blackmun join, dissenting.

The Court’s opinion demeans the importance of the right to counsel. As a practical matter, the opinion probably will have only a slight impact on current custodial interrogation procedures. As a theoretical matter, the Court’s innovative development of an “offense-specific” limitation on the scope of the attorney-client relationship can only generate confusion in the law and undermine the protections that undergird our adversarial system of justice. As a symbolic matter, today’s decision is ominous because it reflects a preference for an inquisitorial system that regards the defense lawyer as an impediment rather than a servant to the cause of justice.

*184r — <

The predicate for the Court’s entire analysis is the failure of the defendant at the preliminary hearing to make a “statement that can reasonably be construed to be expression of a desire for the assistance of an attorney in dealing with custodial interrogation by the police.” Ante, at 178. If petitioner in this case had made such a statement indicating that he was involáng his Fifth Amendment right to counsel as well as his Sixth Amendment right to counsel, the entire offense-specific house of cards that the Court has erected today would collapse, pursuant to our holding in Arizona v. Roberson, 486 U. S. 675 (1988), that a defendant who invokes the right to counsel for interrogation on one offense may not be reapproached regarding any offense unless counsel is present.

In future expected to make sure that they, or their clients, make a statement on the record that will obviate the consequences of today’s holding. That is why I think this decision will have little, if any, practical effect on police practices.

II

The outcome of this case is determined by the Court s parsimonious “offense-specific” description of the right to counsel guaranteed by the Sixth Amendment. The Court’s definition is inconsistent with the high value our prior cases have placed on this right, with the ordinary understanding of the scope of the right, and with the accepted practice of the legal profession.

In Michigan v. Jackson, 475 we the defendant’s invocation of his right to the assistance of counsel at arraignment prohibited the police from initiating a postarraignment custodial interrogation without notice to his lawyer. After explaining that our prior cases required us “to give a broad, rather than a narrow, interpretation to a defendant’s request for counsel,” we squarely rejected “the *185State’s suggestion that respondents’ requests for the appointment of counsel should be construed to apply only to representation in formal legal proceedings.” Id., at 633. Instead, we noted that “it is the State that has the burden of establishing a valid waiver [of the right to counsel]. Doubts must be resolved in favor of protecting the constitutional claim.” Ibid. (citation omitted).

Today, however, the Court accepts a narrow, rather than a broad, interpretation of the same right. It accepts the State’s suggestion that although, under our prior holding in Michigan v. Jackson, a request for the assistance of counsel at a formal proceeding such as an arraignment constitutes an invocation of the right to counsel at police-initiated custodial interrogation as well, such a request only covers interrogation about the specific charge that has already been filed and for which the formal proceeding was held. Today’s approach of construing ambiguous requests for counsel narrowly and presuming a waiver of rights is the opposite of that taken in Jackson.

The Court’s holding today moreover rejects the commonsense evaluation of the nature of an accused’s request for counsel that we expressly endorsed in Jackson:

“We also agree with the comments of the Michigan Supreme Court about the nature of an accused’s request for counsel:

“‘Although judges and lawyers may understand and appreciate the subtle distinctions between the Fifth and Sixth Amendment rights to counsel, the average person does not. When an accused requests an attorney, either before a police officer or a magistrate, he does not know which constitutional right he is invoking; he therefore should not be expected to articulate exactly why or for what purposes he is seeking counsel. It makes little sense to afford relief from further interrogation to a defendant who asks a police officer for an attorney, but permit further interrogation to a defendant who makes *186an identical request to a judge. The simple fact that defendant has requested an attorney indicátes that he does not believe that he is sufficiently capable of dealing with his adversaries singlehandedly.’ 421 Mich., at 63-64, 365 N. W. 2d, at 67.” Id., at 633-634, n. 7.

The Court explains away this commonsense understanding by stating that although “[t]hose observations were perhaps true in the context of deciding whether a request for the assistance of counsel in defending against a particular charge implied a desire to have that counsel serve as an ‘intermediary’ for all further interrogation on that charge[, t]hey are assuredly not true in the quite different context of deciding whether such a request implies a desire never to undergo custodial interrogation, about anything, without counsel present.” Ante, at 180, n. 1. Even assuming that this explanation by the Court could be supported if the custodial interrogation related to an offense that was entirely separate from the charge for which a suspect had invoked his Sixth Amendment right to counsel, it cannot explain away the commonsense reality that petitioner in this case could not have known that his invocation of his Sixth Amendment right to counsel was restricted to the Milwaukee County offense, given that investigations of the Milwaukee County offense and the Caledonia offense were concurrent and conducted by overlapping personnel.1

*187Finally, the Court’s “offense-specific” characterization of the constitutional right to counsel ignores the substance of the attorney-client relationship that the legal profession has developed over the years. The scope of the relationship between an individual accused of crime and his attorney is as broad as the subject matter that might reasonably be encompassed by negotiations for a plea bargain or the contents of a presentence investigation report. Any notion that a constitutional right to counsel is, or should be, narrowly defined by the elements of a pending charge is both unrealistic and invidious. Particularly given the implication that McNeil would be given favorable treatment if he told “his side of the story” as to either or both crimes to the Milwaukee County officers, I find the Court’s restricted construal of McNeil’s relationship with his appointed attorney at the arraignment on the armed robbery charges to be unsupported.

In any case, the offense-specific limitation on the Sixth Amendment right to counsel can only generate confusion in the law. The parties and the Court have assumed in this case, for the purposes of analyzing the legal issues, that the custodial interrogation of McNeil involved an offense (murder) that was completely unrelated to the pending charge of armed robbery. The Court therefore does not flesh out the precise boundaries of its newly created “offense-specific” limitation on a venerable constitutional right. I trust its boundaries will not be patterned after the Court’s double jeopardy jurisprudence, cf. Blockburger v. United States, 284 U. S. 299 (1932), and I can only wonder how much leeway it will accord the police to file charges selectively in order to preserve opportunities for custodial interrogation, particularly if the Court is so unquestioningly willing to treat the offenses in this case as separate even though the investigations were *188concurrent and conducted by overlapping personnel. Whatever the future may portend, the Court’s new rule can only dim the “bright-line” quality of prior cases such as Edwards v. Arizona, 451 U. S. 477 (1981), Solem v. Stumes, 465 U. S. 638 (1984), and Michigan v. Jackson, 475 U. S. 625 (1986).

Ill

In the final analysis, the fear that making counsel available to persons held in custody would “seriously impede effective law enforcement.” Ante, at 180. The magnitude of the Court’s alarm is illuminated by its use of italics:

“Thus, if we were to adopt persons in pretrial custody for serious offenses would be unapproachable by police officers suspecting them of involvement in other crimes, even though they have never expressed any unwillingness to be questioned.” Ante, at 181.

Of course, the Court is quite wrong and its fears are grossly exaggerated. The fears are exaggerated because, as I have explained, today’s holding will probably affect very few cases in the future. The fears are misguided because a contrary rule would not make all pretrial detainees “unapproachable”; it would merely serve to ensure that a suspect’s statements during custodial interrogation are truly voluntary.

A contrary rule would tion. Undergirding our entire line of cases requiring the police to follow fair procedures when they interrogate presumptively innocent citizens suspected of criminal wrongdoing is the longstanding recognition that an adversarial system of justice can function effectively only when the adversaries communicate with one another through counsel and when laypersons are protected from overreaching by more experienced and skilled professionals. Whenever the Court ignores the importance of fair procedure in this context and describes the societal interest in obtaining “uncoerced confes*189sions” from pretrial detainees as an “unmitigated good,” the Court is revealing a preference for an inquisitorial system of justice. As I suggested in Moran v. Burbine, 475 U. S. 412 (1986):

“This case turns on a proper appraisal of the role of the lawyer in our society. If a lawyer is seen as a nettlesome obstacle to the pursuit of wrongdoers — as in an inquisitorial society — then the Court’s decision today makes a good deal of sense. If a lawyer is seen as an aid to the understanding and protection of constitutional rights — as in an accusatorial society — then today’s decision makes no sense at all.” Id., at 468 (Stevens, J., dissenting).

The Court’s refusal to acknowledge any “danger of ‘subtle compulsion’”2 in a case of this kind evidences an inability to recognize the difference between an inquisitorial and an adversarial system of justice. Accordingly, I respectfully dissent.

Justice Scalia

delivered the opinion of the Court.

We consider in this case the scope and continued viability of the rule announced by this Court in Michigan v. Jackson, 475 U. S. 625 (1986), forbidding police to initiate interroga*781tion of a criminal defendant once he has requested counsel at an arraignment or similar proceeding.

I

Petitioner Jesse Montejo was arrested on September 6, 2002, in connection with the robbery and murder of Lewis Ferrari, who had been found dead in his own home one day earlier. Suspicion quickly focused on Jerry Moore, a disgruntled former employee of Ferrari’s dry cleaning business. Police sought to question Montejo, who was a known associate of Moore.

Montejo waived his rights under Miranda v. Arizona, 384 U. S. 436 (1966), and was interrogated at the sheriff’s office by police detectives through the late afternoon and evening of September 6 and the early morning of September 7. During the interrogation, Montejo repeatedly changed his account of the crime, at first claiming that he had only driven Moore to the victim’s home, and ultimately admitting that he had shot and killed Ferrari in the course of a botched burglary. These police interrogations were videotaped.

On September 10, Montejo was brought before a judge for what is known in Louisiana as a “72-hour hearing” — a preliminary hearing required under state law.1 Although the proceedings were not transcribed, the minute record indicates what transpired: “The defendant being charged with First Degree Murder, Court ordered N[o] Bond set in this matter. Further, Court ordered the Office of Indigent Defender be appointed to represent the defendant.” App. to Pet. for Cert. 63a.

Later that same day, two police detectives visited Montejo back at the prison and requested that he accompany them on an excursion to locate the murder weapon (which Montejo *782had earlier indicated he had thrown into a lake). After some back-and-forth, the substance of which remains in dispute, Montejo was again read his Miranda rights and agreed to go along; during the excursion, he wrote an inculpatory letter of apology to the victim’s widow. Only upon their return did Montejo finally meet his court-appointed attorney, who was quite upset that the detectives had interrogated his client in his absence.

At trial, the letter of apology was admitted over defense objection. The jury convicted Montejo of first-degree murder, and he was sentenced to death.

The Louisiana Supreme Court affirmed the conviction and sentence. 06-1807 (1/16/08), 974 So. 2d 1238 (2008). As relevant here, the court rejected Monte jo’s argument that under the rule of Jackson, supra, the letter should have been suppressed. 974 So. 2d, at 1261. Jackson held that “if police initiate interrogation after a defendant’s assertion, at an arraignment or similar proceeding, of his right to counsel, any waiver of the defendant’s right to counsel for that police-initiated interrogation is invalid.” 475 U. S., at 636.

Citing a decision of the United States Court of Appeals for the Fifth Circuit, Montoya v. Collins, 955 F. 2d 279 (1992), the Louisiana Supreme Court reasoned that the prophylactic protection of Jackson is not triggered unless and until the defendant has actually requested a lawyer or has otherwise asserted his Sixth Amendment right to counsel. 974 So. 2d, at 1260-1261, and n. 68. Because Montejo simply stood mute at his 72-hour hearing while the judge ordered the appointment of counsel, he had made no such request or assertion. So the proper inquiry, the court ruled, was only whether he had knowingly, intelligently, and voluntarily waived his right to have counsel present during the interaction with the police. Id., at 1261. And because Montejo had been read his Miranda rights and agreed to waive them, *783the Court answered that question in the affirmative, 974 So. 2d, at 1262, and upheld the conviction.

We granted certiorari. 554 U. S. 944 (2008).

II

Montejo and his amici raise a number of pragmatic objections to the Louisiana Supreme Court’s interpretation of Jackson. We agree that the approach taken below would lead either to an unworkable standard, or to arbitrary and anomalous distinctions between defendants in different States. Neither would be acceptable.

Under the rule adopted by the Louisiana Supreme Court, a criminal defendant must request counsel, or otherwise “assert” his Sixth Amendment right at the preliminary hearing, before the Jackson protections are triggered. If he does so, the police may not initiate further interrogation in the absence of counsel. But if the court on its own appoints counsel, with the defendant taking no affirmative action to invoke his right to counsel, then police are free to initiate further interrogations provided that they first obtain an otherwise valid waiver by the defendant of his right to have counsel present.

This rule would apply well enough in States that require the indigent defendant formally to request counsel before any appointment is made, which usually occurs after the court has informed him that he will receive counsel if he asks for it. That is how the system works in Michigan, for example, Mich. Ct. Rule 6.005(A) (2009), whose scheme produced the factual background for this Court’s decision in Michigan v. Jackson. Jackson, like all other represented indigent defendants in the State, had requested counsel in accordance with the applicable state law.

But many States follow other practices. In some two dozen, the appointment of counsel is automatic upon a finding of indigency, e. g., Kan. Stat. Ann. §22-4503(c) (2007); and in *784a number of others, appointment can be made either upon the defendant’s request or sua sponte by the court, e. g., Del. Code Ann., Tit. 29, § 4602(a) (2003). See App. to Brief for National Legal Aid & Defender Assn, et al. as Amici Curiae la-21a. Nothing in our Jackson opinion indicates whether we were then aware that not all States require that a defendant affirmatively request counsel before one is appointed; and of course we had no occasion there to decide how the rule we announced would apply to these other States.

The Louisiana Supreme Court’s answer to that unresolved question is troublesome. The central distinction it draws— between defendants who “assert” their right to counsel and those who do not — is exceedingly hazy when applied to States that appoint counsel absent request from the defendant. How to categorize a defendant who merely asks, prior to appointment, whether he will be appointed counsel? Or who inquires, after the fact, whether he has been? What treatment for one who thanks the court after the appointment is made? And if the court asks a defendant whether he would object to appointment, will a quick shake of his head count as an assertion of his right?

To the extent that the Louisiana Supreme Court’s rule also permits a defendant to trigger Jackson through the “acceptance” of counsel, that notion is even more mysterious: How does one affirmatively accept counsel appointed by court order? An indigent defendant has no right to choose his counsel, United States v. Gonzalez-Lopez, 548 U. S. 140, 151 (2006), so it is hard to imagine what his “acceptance” would look like, beyond the passive silence that Monte jo exhibited.

In practice, judicial application of the Louisiana rule in States that do not require a defendant to make a request for counsel could take either of two paths. Courts might ask on a case-by-case basis whether a defendant has somehow invoked his right to counsel, looking to his conduct at the preliminary hearing — his statements and gestures — and the to*785tality of the circumstances. Or, courts might simply determine as a categorical matter that defendants in these States — over half of those in the Union — simply have no opportunity to assert their right to counsel at the hearing and are therefore out of luck.

Neither approach is desirable. The former would be particularly impractical in light of the fact that, as amici describe, preliminary hearings are often rushed, and are frequently not recorded or transcribed. Brief for National Legal Aid & Defender Assn, et al. 25-30. The sheer volume of indigent defendants, see id., at 29, would render the monitoring of each particular defendant’s reaction to the appointment of counsel almost impossible. And sometimes the defendant is not even present. E. g., La. Code Crim. Proc. Ann., Art. 230.1(A) (West Supp. 2009) (allowing court to appoint counsel if defendant is “unable to appear”). Police who did not attend the hearing would have no way to know whether they could approach a particular defendant; and for a court to adjudicate that question ex post would be a fact-intensive and burdensome task, even if monitoring were possible and transcription available. Because “clarity of. .. command” and “certainty of. . . application” are crucial in rules that govern law enforcement, Minnick v. Mississippi, 498 U. S. 146, 151 (1990), this would be an unfortunate way to proceed. See also Moran v. Burbine, 475 U. S. 412, 425-426 (1986).

The second possible course fares no better, for it would achieve clarity and certainty only at the expense of introducing arbitrary distinctions: Defendants in States that automatically appoint counsel would have no opportunity to invoke their rights and trigger Jackson, while those in other States, effectively instructed by the court to request counsel, would be lucky winners. That sort of hollow formalism is out of place in a doctrine that purports to serve as a practical safeguard for defendants’ rights.

*786III

But if the Louisiana Supreme Court’s application of Jackson is unsound as a practical matter, then Montejo’s solution is untenable as a theoretical and doctrinal matter. Under his approach, once a defendant is represented by counsel, police may not initiate any further interrogation. Such a rule would be entirely untethered from the original rationale of Jackson.

A

It is worth emphasizing first what is not in dispute or at stake here. Under our precedents, once the adversary judicial process has been initiated, the Sixth Amendment guarantees a defendant the right to have counsel present at all “critical” stages of the criminal proceedings. United States v. Wade, 388 U. S. 218, 227-228 (1967); Powell v. Alabama, 287 U. S. 45, 57 (1932). Interrogation by the State is such a stage. Massiah v. United States, 377 U. S. 201, 204-205 (1964); see also United States v. Henry, 447 U. S. 264, 274 (1980).

Our precedents also place beyond doubt that the Sixth Amendment right to counsel may be waived by a defendant, so long as relinquishment of the right is voluntary, knowing, and intelligent. Patterson v. Illinois, 487 U. S. 285,292, n. 4 (1988); Brewer v. Williams, 430 U. S. 387, 404 (1977); Johnson v. Zerbst, 304 U. S. 458, 464 (1938). The defendant may waive the right whether or not he is already represented by counsel; the decision to waive need not itself be counseled. Michigan v. Harvey, 494 U. S. 344, 352-353 (1990). And when a defendant is read his Miranda rights (which include the right to have counsel present during interrogation) and agrees to waive those rights, that typically does the trick, even though the Miranda rights purportedly have their source in the Fifth Amendment:

“As a general matter ... an accused who is admonished with the warnings prescribed by this Court in Miranda *787. . . has been sufficiently apprised of the nature of his Sixth Amendment rights, and of the consequences of abandoning those rights, so that his waiver on this basis will be considered a knowing and intelligent one.” Patterson, supra, at 296.

The only question raised by this case, and the only one addressed by the Jackson rule, is whether courts must presume that such a waiver is invalid under certain circumstances. 475 U. S., at 630, 633. We created such a presumption in Jackson by analogy to a similar prophylactic rule established to protect the Fifth Amendment-based Miranda right to have counsel present at any custodial interrogation. Edwards v. Arizona, 451 U. S. 477 (1981), decided that once “an accused has invoked his right to have counsel present during custodial interrogation ... [he] is not subject to further interrogation by the authorities until counsel has been made available,” unless he initiates the contact. Id., at 484-485.

The Edwards rule is “designed to prevent police from badgering a defendant into waiving his previously asserted Miranda rights,” Harvey, supra, at 350. It does this by presuming his postassertion statements to be involuntary, “even where the suspect executes a waiver and his statements would be considered voluntary under traditional standards.” McNeil v. Wisconsin, 501 U. S. 171, 177 (1991). This prophylactic rule thus “protects] a suspect’s voluntary choice not to speak outside his lawyer’s presence.” Texas v. Cobb, 532 U. S. 162, 175 (2001) (Kennedy, J., concurring).

Jackson represented a “wholesale importation of the Edwards rule into the Sixth Amendment.” Cobb, supra, at 175. The Jackson Court decided that a request for counsel at an arraignment should be treated as an invocation of the Sixth Amendment right to counsel “at every critical stage of the prosecution,” 475 U. S., at 633, despite doubt that defendants “actually inten[d] their request for counsel to encompass representation during any further questioning,” id., at 632-*788633, because doubts must be “resolved in favor of protecting the constitutional claim,” id., at 633. Citing Edwards, the Court held that any subsequent waiver would thus be “insufficient to justify police-initiated interrogation.” 475 U. S., at 635. In other words, we presume such waivers involuntary “based on the supposition that suspects who assert their right to counsel are unlikely to waive that right voluntarily” in subsequent interactions with police. Harvey, supra, at 350.

In his dissent, Justice Stevens presents us with a revisionist view of Jackson. The defendants' request for counsel, he contends, was important only because it proved that counsel had been appointed. Such a non sequitur (nowhere alluded to in the case) hardly needs rebuttal. Proceeding from this fanciful premise, he claims that the decision actually established “a rule designed to safeguard a defendant's right to rely on the assistance of counsel,” post, at 807 (hereinafter dissent), not one “designed to prevent police badgering,” ibid. To safeguard the right to assistance of counsel from what? From a knowing and voluntary waiver by the defendant himself? Unless the dissent seeks to prevent a defendant altogether from waiving his Sixth Amendment rights, i. e., to “imprison a man in his privileges and call it the Constitution,” Adams v. United States ex rel. McCann, 317 U. S. 269, 280 (1942) — a view with zero support in reason, history, or case law — the answer must be: from police pressure, i. e., badgering. The antibadgering rationale is the only way to make sense of Jackson’s repeated citations of Edwards, and the only way to reconcile the opinion with our waiver jurisprudence.2

*789B

With this understanding of what Jackson stands for and whence it came, it should be clear that Montejo’s interpretation of that decision — that no represented defendant can ever be approached by the State and asked to consent to interrogation — is off the mark. When a court appoints counsel for an indigent defendant in the absence of any request on his part, there is no basis for a presumption that any subsequent waiver of the right to counsel will be involuntary. There is no “initial election” to exercise the right, Patterson, 487 U. S., at 291, that must be preserved through a prophylactic rule against later waivers. No reason exists to assume that a defendant like Montejo, who has done nothing at all to express his intentions with respect to his Sixth Amendment rights, would not be perfectly amenable to speaking with the police without having counsel present. And no reason exists to prohibit the police from inquiring. Edwards and Jackson are meant to prevent police from badgering defendants into changing their minds about their rights, but a defendant who never asked for counsel has not yet made up his mind in the first instance.

The dissent’s argument to the contrary rests on a flawed a fortiori: “If a defendant is entitled to protection from police-initiated interrogation under the Sixth Amendment when he merely requests a lawyer, he is even more obviously entitled to such protection when he has secured a lawyer.” Post, at 804. The question in Jackson, however, was not whether respondents were entitled to counsel (they unquestionably were), but “whether respondents validly waived their right to counsel,” 475 U. S., at 630; and even if it is reasonable to presume from a defendant's request for counsel that any subsequent waiver of the right was coerced, no such *790presumption can seriously be entertained when a lawyer was merely “secured” on the defendant’s behalf, by the State itself, as a matter of course. Of course, reading the dissent’s analysis, one would have no idea that Montejo executed any waiver at all.

In practice, Montejo’s rule would prevent police-initiated interrogation entirely once the Sixth Amendment right attaches, at least in those States that appoint counsel promptly without request from the defendant. As the dissent in Jackson pointed out, with no expressed disagreement from the majority, the opinion “most assuredly [did] not hold that the Edwards per se rule prohibiting all police-initiated interrogations applies from the moment the defendant’s Sixth Amendment right to counsel attaches, with or without a request for counsel by the defendant.” 475 U. S., at 640 (opinion of Rehnquist, J.). That would have constituted a “shockingly dramatic restructuring of the balance this Court has traditionally struck between the rights of the defendant and those of the larger society.” Ibid.

Montejo’s rule appears to have its theoretical roots in codes of legal ethics, not the Sixth Amendment. The American Bar Association’s Model Rules of Professional Conduct (which nearly all States have adopted into law in whole or in part) mandate that “a lawyer shall not communicate about the subject of [a] representation with a person the lawyer knows to be represented by another lawyer in the matter, unless the lawyer has the consent of the other lawyer or is authorized to do so by law or a court order.” Model Rule 4.2 (2008). But the Constitution does not codify the ABA’s Model Rules, and does not make investigating police officers lawyers. Montejo’s proposed rule is both broader and narrower than the Model Rule. Broader, because Montejo would apply it to all agents of the State, including the detectives who interrogated him, while the ethical rule governs only lawyers. And narrower, because he agrees that if a defendant initiates contact with the police, they may talk *791freely — whereas a lawyer could be sanctioned for interviewing a represented party even if that party “initiates” the communication and consents to the interview. Model Rule 4.2, Comment 3.

Monte jo contends that our decisions support his interpretation of the Jackson rule. We think not. Many of the eases he cites concern the substantive scope of the Sixth Amendment — e. g., whether a particular interaction with the State constitutes a “critical” stage at which counsel is entitled to be present — not the validity of a Sixth Amendment waiver. See Maine v. Moulton, 474 U. S. 159 (1985); Henry, 447 U. S. 264; Massiah, 377 U. S. 201; see also Moran, 475 U. S. 412. Since everyone agrees that absent a valid waiver, Monte jo was entitled to a lawyer during the interrogation, those cases do not advance his argument.

Montejo also points to descriptions of the Jackson holding in two later cases. In one, we noted that “analysis of the waiver issue changes” once a defendant “obtains or even requests counsel.” Harvey, 494 U. S., at 352. But elsewhere in the same opinion, we explained that Jackson applies “after a defendant requests assistance of counsel,” 494 U. S., at 349; “when a suspect charged with a crime requests counsel outside the' context of interrogation,” id., at 350; and to “suspects who assert their right to counsel,” ibid. The accuracy of the “obtains” language is thus questionable. Anyway, since Harvey held that evidence obtained in violation of the Jackson rule could be admitted to impeach the defendant’s trial testimony, 494 U. S., at 346, the Court’s varying descriptions of when the rule was violated were dicta. The dictum from the other decision, Patterson, supra, at 290, n. 3, is no more probative.3

*792The upshot is that even on Jackson’s own terms, it would be completely unjustified to presume that a defendant’s consent to police-initiated interrogation was involuntary or coerced simply because he had previously been appointed a lawyer.

IV

So on the one hand, requiring an initial “invocation” of the right to counsel in order to trigger the Jackson presumption is consistent with the theory of that decision, but (as Montejo and his amici argue, see Part II, swpra) would be unworkable in more than half the States of the Union. On the other hand, eliminating the invocation requirement would render the rule easy to apply but depart fundamentally from the Jackson rationale.

We do not think that stare decisis requires us to expand significantly the holding of a prior decision — fundamentally revising its theoretical basis in the process — in order to cure its practical deficiencies. To the contrary, the fact that a decision has proved “unworkable” is a traditional ground for overruling it. Payne v. Tennessee, 501 U. S. 808, 827 (1991). Accordingly, we called for supplemental briefing addressed to the question whether Michigan v. Jackson should be overruled.

Beyond workability, the relevant factors in deciding whether to adhere to the principle of stare decisis include the antiquity of the precedent, the reliance interests at stake, *793and of course whether the decision was well reasoned. Pearson v. Callahan, 555 U. S. 223, 234-235 (2009). The first two cut in favor of abandoning Jackson: The opinion is only two decades old, and eliminating it would not upset expectations. Any criminal defendant learned enough to order his affairs based on the rule announced in Jackson would also be perfectly capable of interacting with the police on his own. Of course it is likely true that police and prosecutors have been trained to comply with Jackson, see generally Supplemental Brief for Larry D. Thompson et al. as Amici Curiae, but that is hardly a basis for retaining it as a constitutional requirement. If a State wishes to abstain from requesting interviews with represented defendants when counsel is not present, it obviously may continue to do so.4

Which brings us to the strength of Jackson’s reasoning. When this Court creates a prophylactic rule in order to protect a constitutional right, the relevant “reasoning” is the weighing of the rule’s benefits against its costs. “The value of any prophylactic rule ... must be assessed not only on the basis of what is gained, but also on the basis of what is lost.” Minnick, 498 U. S., at 161 (Scalia, J., dissenting). We think that the marginal benefits of Jackson (viz., the number of confessions obtained coercively that are suppressed by its bright-line rule and would otherwise have been admitted) are dwarfed by its substantial costs (viz., hindering “society’s compelling interest in finding, convicting, and punishing those who violate the law,” Moran, supra, at 426).

*794What does the Jackson rule actually achieve by way of preventing unconstitutional conduct? Recall that the purpose of the rule is to preclude the State from badgering defendants into waiving their previously asserted rights. See Harvey, 494 U. S., at 350; see also McNeil, 501 U. S., at 177. The effect of this badgering might be to coerce a waiver, which would render the subsequent interrogation a violation of the Sixth Amendment. See Massiah, 377 U. S., at 204. Even though involuntary waivers are invalid even apart from Jackson, see Patterson, 487 U. S., at 292, n. 4, mistakes are of course possible when courts conduct case-by-case voluntariness review. A bright-line rule like that adopted in Jackson ensures that no fruits of interrogations made possible by badgering-induced involuntary waivers are ever erroneously admitted at trial.

But without Jackson, how many would be? The answer is few if any. The principal reason is that the Court has already taken substantial other, overlapping measures toward the same end. Under Miranda's prophylactic protection of the right against compelled self-incrimination, any suspect subject to custodial interrogation has the right to have a lawyer present if he so requests, and to be advised of that right. 384 U. S., at 474. Under Edwards' prophylactic protection of the Miranda right, once such a defendant “has invoked his right to have counsel present,” interrogation must stop. 451 U. S., at 484. And under Minnick’s prophylactic protection of the Edwards right, no subsequent interrogation may take place until counsel is present, “whether or not the accused has consulted with his attorney.” 498 U. S., at 153.

These three layers of prophylaxis are sufficient. Under the Miranda-Edwards-Minnick line of eases (which is not in doubt), a defendant who does not want to speak to the police without counsel present need only say as much when he is first approached and given the Miranda warnings. At that point, not only must the immediate contact end, but *795“badgering” by later requests is prohibited. If that regime suffices to protect the integrity of “a suspect’s voluntary choice not to speak outside his lawyer’s presence” before his arraignment, Cobb, 532 U. S., at 175 (Kennedy, J., concurring), it is hard to see why it would not also suffice to protect that same choice after arraignment, when Sixth Amendment rights have attached. And if so, then Jackson is simply superfluous.

It is true, as Montejo points out in his supplemental brief, that the doctrine established by Miranda and Edwards is designed to protect Fifth Amendment, not Sixth Amendment, rights. But that is irrelevant. What matters is that these cases, like Jackson, protect the right to have counsel during custodial interrogation — which right happens to be guaranteed (once the adversary judicial process has begun) by two sources of law. Since the right under both sources is waived using the same procedure, Patterson, supra, at 296, doctrines ensuring voluntariness of the Fifth Amendment waiver simultaneously ensure the voluntariness of the Sixth Amendment waiver.

Montejo also correctly observes that the Miranda-Edwards regime is narrower than Jackson in one respect: The former applies only in the context of custodial interrogation. If the defendant is not in custody then those decisions do not apply; nor do they govern other, noninterrogative types of interactions between the defendant and the State (like pretrial lineups). However, those uncovered situations are the least likely to pose a risk of coerced waivers. When a defendant is not in custody, he is in control, and need only shut his door or walk away to avoid police badgering. And noninterrogative interactions with the State do not involve the “inherently compelling pressures,” Miranda, supra, at 467, that one might reasonably fear could lead to involuntary waivers.

Jackson was policy driven, and if that policy is being adequately served through other means, there is no reason to *796retain its rule. Miranda and the cases that elaborate upon it already guarantee not simply noncoercion in the traditional sense, but what Justice Harlan referred to as “voluntariness with a vengeance,” 384 U. S., at 505 (dissenting opinion). There is no need to take Jackson’s further step of requiring voluntariness on stilts.

On the other side of the equation are the costs of adding the bright-line Jackson rule on top of Edwards and other extant protections. The principal cost of applying any exclusionary rule “is, of course, letting guilty and possibly dangerous criminals go free . . . .” Herring v. United States, 555 U. S. 135, 141 (2009). Jackson not only “operates to invalidate a confession given by the free choice of suspects who have received proper advice of their Miranda rights but waived them nonetheless,” Cobb, supra, at 174-175 (Kennedy, J., concurring), but also deters law enforcement officers from even trying to obtain voluntary confessions. The “ready ability to obtain uncoerced confessions is not an evil but an unmitigated good.” McNeil, supra, at 181. Without these confessions, crimes go unsolved and criminals unpunished. These are not negligible costs, and in our view the Jackson Court gave them too short shrift.5

Notwithstanding this calculus, Montejo and his amici urge the retention of Jackson. Their principal objection to its elimination is that the Edwards regime which remains will not provide an administrable rule. But this Court has praised Edwards precisely because it provides “‘clear and unequivocal’ guidelines to the law enforcement profession,” Arizona v. Roberson, 486 U. S. 675, 682 (1988). Our cases *797make clear which sorts of statements trigger its protections, see Davis v. United States, 512 U. S. 452, 459 (1994), and once triggered, the rule operates as a bright line. Monte jo expresses concern that courts will have to determine whether statements made at preliminary hearings constitute Edwards invocations — thus implicating all the practical problems of the Louisiana rule we discussed above, see Part II, supra. That concern is misguided. “We have in fact never held that a person can invoke his Miranda rights anticipatorily, in a context other than ‘custodial interrogation’. . . .” McNeil, 501 U. S., at 182, n. 3. What matters for Miranda and Edwards is what happens when the defendant is approached for interrogation, and (if he consents) what happens during the interrogation — not what happened at any preliminary hearing.

In sum, when the marginal benefits of the Jackson rule are weighed against its substantial costs to the truth-seeking process and the criminal justice system, we readily conclude that the rule does not “pay its way,” United States v. Leon, 468 U. S. 897, 907-908, n. 6 (1984). Michigan v. Jackson should be and now is overruled.

Y

Although our holding means that the Louisiana Supreme Court correctly rejected Montejo’s claim under Jackson, we think that Monte jo should be given an opportunity to contend that his letter of apology should still have been suppressed under the rule of Edwards. If Monte jo made a clear assertion of the right to counsel when the officers approached him about accompanying them on the excursion for the murder weapon, then no interrogation should have taken place unless Monte jo initiated it. Davis, supra, at 459. Even if Montejo subsequently agreed to waive his rights, that waiver would have been invalid had it followed an “unequivocal election of the right,” Cobb, 532 U. S., at 176 (Kennedy, J., concurring).

*798Montejo understandably did not pursue an Edwards objection, because Jackson served as the Sixth Amendment analogy to Edwards and offered broader protections. Our decision today, overruling Jackson, changes the legal landscape and does so in part based on the protections already provided by Edwards. Thus we think that a remand is appropriate so that Montejo can pursue this alternative avenue for relief. Montejo may also seek on remand to press any claim he might have that his Sixth Amendment waiver was not knowing and voluntary, e. g., his argument that the waiver was invalid because it was based on misrepresentations by police as to whether he had been appointed a lawyer, cf. Moran, 475 U. S., at 428-429. These matters have heightened importance in light of our opinion today.

We do not venture to resolve these issues ourselves, not only because we are a court of final review, “not of first view,” Cutter v. Wilkinson, 544 U. S. 709, 718, n. 7 (2005), but also because the relevant facts remain unclear. Montejo and the police gave inconsistent testimony about exactly what took place on the afternoon of September 10, 2002, and the Louisiana Supreme Court did not make an explicit credibility determination. Moreover, Monte jo’s testimony came not at the suppression hearing, but rather only at trial, and we are unsure whether under state law that testimony came too late to affect the propriety of the admission of the evidence. These matters are best left for resolution on remand.

We do reject, however, the dissent’s revisionist legal analysis of the “knowing and voluntary” issue. Post, at 810-814. In determining whether a Sixth Amendment waiver was knowing and voluntary, there is no reason categorically to distinguish an unrepresented defendant from a represented one. It is equally true for each that, as we held in Patterson, the Miranda warnings adequately inform him “of his right to have counsel present during the questioning,” and make him “aware of the consequences of a decision by him *799to waive his Sixth Amendment rights,” 487 U. S., at 293. Somewhat surprisingly for an opinion that extols the virtues of stare decisis, the dissent complains that our “treatment of the waiver question rests entirely on the dubious decision in Patterson,” post, at 812. The Court in Patterson did not consider the result dubious, nor does the Court today.

* * *

This case is an exemplar of Justice Jackson’s oft quoted warning that this Court “is forever adding new stories to the temples of constitutional law, and the temples have a way of collapsing when one story too many is added.” Douglas v. City of Jeannette, 319 U. S. 157, 181 (1943) (opinion concurring in result). We today remove Michigan v. Jackson’s fourth story of prophylaxis.

The judgment of. the Louisiana Supreme Court is vacated, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

Justice Alito, with whom Justice Kennedy joins, concurring.

Earlier this Term, in Arizona v. Gant, ante, p. 332, the Court overruled New York v. Belton, 453 U. S. 454 (1981), even though that case had been on the books for 28 years, had not been undermined by subsequent decisions, had been recently reaffirmed and extended, had proved to be eminently workable (indeed, had been adopted for precisely that reason), and had engendered substantial law enforcement reliance. See Gant, ante, at 358 (Alito, J., dissenting). The Court took this step even though we were not asked to overrule Belton, and this new rule is almost certain to lead to a host of problems. See Gant, ante, at 363-365 (Alito, J., dissenting); Megginson v. United States, post, p. 1230; Grooms v. United States, post, p. 1231 (same).

*800Justice Scalia,

who cast the deciding vote to overrule Belton, dismissed stare decisis concerns with the following observation: “[I]t seems to me ample reason that the precedent was badly reasoned and produces erroneous ... results.” Gant, ante, at 353 (concurring opinion). This narrow view of stare decisis provides the only principle on which the decision in Gant can be justified.

In light of Gant, the discussion of stare decisis in Justice Stevens’ dissent* is surprising. His dissent in the case at hand criticizes the Court for “[a]cting on its own” in reconsidering Michigan v. Jackson, 475 U. S. 625 (1986). Post, at 804 (hereinafter dissent). But the same was true in Gant, and in this case, the Court gave the parties and interested amici the opportunity to submit supplemental briefs on the issue, a step not taken in Gant.

The dissent faults the Court for “cast[ing] aside the reliance interests of law enforcement,” post, at 809, but in Gant, there were real and important law enforcement interests at stake, see ante, at 358-360 (Alito, J., dissenting). Even the Court conceded that the Belton rule had “been widely taught in police academies and that law enforcement officers ha[d] relied on the rule in conducting vehicle searches during the past 28 years.” Ante, at 349. And whatever else might be said about Belton, it surely provided a bright-line rule.

A month ago, none of this counted for much, but today the dissent writes:

“Jackson’s bright-line rule has provided law enforcement officers with clear guidance, allowed prosecutors to quickly and easily assess whether confessions will be admissible in court, and assisted judges in determining whether a defendant’s Sixth Amendment rights have been violated by police interrogation.” Post, at 808.

*801It is striking that precisely the same points were true in Gant:

“[Belton’s] bright-line rule ha[d] provided law enforcement officers with clear guidance, allowed prosecutors to quickly and easily assess whether [evidence obtained in a vehicle search] w[ould] be admissible in court, and assisted judges in determining whether a defendant’s [Fourth] Amendment rights ha[d] been violated by police interrogation.” Post, at 808.

The dissent, finally, invokes Jackson’s antiquity, stating that “the 23-year existence of a simple bright-line rule” should weigh in favor of its retention. Post, at 810. But in Gant, the Court had no compunction about casting aside a 28-year-old bright-line rule. I can only assume that the dissent thinks that our constitutional precedents are like certain wines, which are most treasured when they are neither too young nor too old, and that Jackson, supra, at 23, is in its prime, whereas Belton, supra, at 28, had turned brownish and vinegary.

I agree with the dissent that stare decisis should promote “ ‘the evenhanded ... development of legal principles,’ ” post, at 807 (quoting Payne v. Tennessee, 501 U. S. 808, 827-828 (1991)). The treatment of stare decisis in Gant fully supports the decision in the present case.

Justice Stevens, with whom Justice Souter and Justice Ginsburg join, and with whom Justice Breyer joins except for footnote 5, dissenting.

Today the Court properly concludes that the Louisiana Supreme Court’s parsimonious reading of our decision in Michigan v. Jackson, 475 U. S. 625 (1986), is indefensible. Yet the Court does not reverse. Rather, on its own initiative and without any evidence that the longstanding Sixth Amendment protections established in Jackson have caused any harm to the workings of the criminal justice system, the *802Court rejects Jackson outright on the ground that it is “untenable as a theoretical and doctrinal matter.” Ante, at 786. That conclusion rests on a misinterpretation of Jackson’s rationale and a gross undervaluation of the rule of stare decisis. The police interrogation in this case clearly violated petitioner’s Sixth Amendment right to counsel.

I

The Sixth Amendment provides that “[i]n all criminal prosecutions, the accused shall enjoy the right ... to have the Assistance of Counsel for his defence.” The right to counsel attaches during “the initiation of adversary judicial criminal proceedings,” Rothgery v. Gillespie County, 554 U. S. 191, 198 (2008) (internal quotation marks omitted), and it guarantees the assistance of counsel not only during in-court proceedings but during all critical stages, including postarraignment interviews with law enforcement officers, see Patterson v. Illinois, 487 U. S. 285, 290 (1988).

In Jackson, this Court considered whether the Sixth Amendment bars police from interrogating defendants who have requested the appointment of counsel at arraignment. Applying the presumption that such a request constitutes an invocation of the right to counsel “at every critical stage of the prosecution,” 475 U. S., at 633, we held that “a defendant who has been formally charged with a crime and who has requested appointment of counsel at his arraignment” cannot be subject to uncounseled interrogation unless he initiates “exchanges or conversations with the police,” id., at 626.

In this case, petitioner Jesse Montejo contends that police violated his Sixth Amendment right to counsel by interrogating him following his “72-hour hearing” outside the presence of, and without prior notice to, his lawyer. Brief for Petitioner 7. The Louisiana Supreme Court rejected Montejo’s claim. Relying on the fact that the defendants in Jackson had “requested” counsel at arraignment, the state court held that Jackson’s protections did not apply to Mon*803tejo because his counsel was appointed automatically; Montejo had not explicitly requested counsel or affirmatively accepted the counsel appointed to represent him before he submitted to police interrogation. 06-1807, pp. 28-29 (1/16/ 08), 974 So. 2d 1238, 1261.

I agree with the majority’s conclusion that the Louisiana Supreme Court’s decision, if allowed to stand, “would lead either to an unworkable standard, or to arbitrary and anomalous distinctions between defendants in different States,” ante, at 783. Neither option is tolerable, and neither is compelled by Jackson itself.

Our decision in Jackson involved two consolidated cases, both arising in the State of Michigan. Under Michigan law in effect at that time, when a defendant appeared for arraignment the court was required to inform him that counsel would be provided if he was financially needy and he requested representation. Mich. Gen. Ct. Rule 785.4(1) (1976). It was undisputed that the Jackson defendants made such a “request” at their arraignment: one by completing an affidavit of indigency, and the other by responding affirmatively to a question posed to him by the court. See App. in Michigan v. Jackson, O. T. 1984, No. 84-1531, p. 168; App. in Michigan v. Bladel, O. T. 1984, No. 84-1539, pp. 3a-4a. In neither case, however, was it clear that counsel had actually been appointed at the arraignment. Thus, the defendants’ requests for counsel were significant as a matter of state law because they served as evidence that the appointment of counsel had been effectuated even in the absence of proof that defense counsel had actual notice of the appointments.

Unlike Michigan, Louisiana does not require a defendant to make a request in order to receive court-appointed counsel. Consequently, there is no reason to place constitutional significance on the fact that Montejo neither voiced a request for counsel nor affirmatively embraced that appointment post hoc. Certainly our decision in Jackson did not mandate such an odd rule. See ante, at 784 (acknowledging that we *804had no occasion to decide in Jackson how its rule would apply in States that do not make appointment of counsel contingent on affirmative request). If a defendant is entitled to protection from police-initiated interrogation under the Sixth Amendment when he merely requests a lawyer, he is even more obviously entitled to such protection when he has secured a lawyer. Indeed, we have already recognized as much. See Michigan v. Harvey, 494 U. S. 344, 352 (1990) (acknowledging that “once a defendant obtains or even requests counsel,” Jackson alters the waiver analysis); Patterson, 487 U. S., at 290, n. 3 (noting “as a matter of some significance” to the constitutional analysis that defendant had “not retained, or accepted by appointment, a lawyer to represent him at the time he was questioned by authorities” (emphasis added)).1 Once an attorney-client relationship has been established through the appointment or retention of counsel, as a matter of federal law the method by which the relationship was created is irrelevant: The existence of a valid attorney-client relationship provides a defendant with the full constitutional protection afforded by the Sixth Amendment.

II

Today the Court correctly concludes that the Louisiana Supreme Court’s holding is “troublesome,” ante, at 784, “impractical,” ante, at 785, and “unsound,” ante, at 786. Instead of reversing the decision of the state court by simply answering the question on which we granted certiorari in a unanimous opinion, however, the majority has decided to change the law. Acting on its own initiative, the majority overrules Jackson to correct a “theoretical and doctrinal” *805problem of its own imagining, see ante, at 786. A more careful reading of Jackson and the Sixth Amendment cases upon which it relied reveals that the rule announced in Jackson protects a fundamental right that the Court now dishonors.

The majority’s decision to overrule Jackson rests on its assumption that Jackson’s protective rule was intended to “prevent police from badgering defendants into changing their minds about their rights,” ante, at 789; see also ante, at 794, just as the rule adopted in Edwards v. Arizona, 451 U. S. 477 (1981), was designed to prevent police from coercing unindicted suspects into revoking their requests for counsel at interrogation. Operating on that limited understanding of the purpose behind Jackson’s protective rule, the Court concludes that Jackson provides no safeguard not already secured by this Court’s Fifth Amendment jurisprudence. See Miranda v. Arizona, 384 U. S. 436 (1966) (requiring defendants to be admonished of their right to counsel prior to custodial interrogation); Edwards, 451 U. S. 477 (prohibiting police-initiated interrogation following defendant’s invocation of the right to counsel).

The majority’s analysis flagrantly misrepresents Jackson’s underlying rationale and the constitutional interests the decision sought to protect. While it is true that the rule adopted in Jackson was patterned after the rule in Edwards, 451 U. S., at 484-485, the Jackson opinion does not even mention the antibadgering considerations that provide the basis for the Court's decision today. Instead, Jackson relied primarily on cases discussing the broad protections guaranteed by the Sixth Amendment right to counsel — not its Fifth Amendment counterpart. Jackson emphasized that the purpose of the Sixth Amendment is to “ ‘protec[t] the unaided layman at critical confrontations with his adversary,’” 475 U. S., at 631 (quoting United States v. Gouveia, 467 U. S. 180, 189 (1984)), by giving him “ ‘the right to rely on counsel as a “medium” between him[self] and the State,’ ” 475 U. S., at 632 *806(quoting Maine v. Moulton, 474 U. S. 159, 176 (1985)). Underscoring that the commencement of criminal proceedings is a decisive event that transforms a suspect into an accused within the meaning of the Sixth Amendment, we concluded that arraigned defendants are entitled to “at least as much protection” during interrogation as the Fifth Amendment affords unindicted suspects. See, e. g., 475 U. S., at 632 (“[T]he difference between the legal basis for the rule applied in Edwards and the Sixth Amendment claim asserted in these cases actually provides additional support for the application of the rule in these circumstances” (emphasis added)). Thus, although the rules adopted in Edwards and Jackson are similar, Jackson did not rely on the reasoning of Edwards but remained firmly rooted in the unique protections afforded to the attorney-client relationship by the Sixth Amendment.2

Once Jackson is placed in its proper Sixth Amendment context, the majority’s justifications for overruling the decision crumble. Ordinarily, this Court is hesitant to disturb past precedent and will do so only when a rule has proven “outdated, ill-founded, unworkable, or otherwise legitimately *807vulnerable to serious reconsideration.” Vasquez v. Hillery, 474 U. S. 254, 266 (1986). While stare decisis is not “an inexorable command,” we adhere to it as “the preferred course because it promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.” Payne v. Tennessee, 501 U. S. 808, 827-828 (1991).

Paying lipservice to the rule of stare decisis, the majority acknowledges that the Court must consider many factors before taking the dramatic step of overruling a past decision. See ante, at 792-793. Specifically, the majority focuses on four considerations: the reasoning of the decision, the workability of the rule, the reliance interests at stake, and the antiquity of the precedent. The Court exaggerates the considerations favoring reversal, however, and gives short shrift to the valid considerations favoring retention of the Jackson rule.

First, and most central to the Court’s decision to overrule Jackson, is its assertion that Jackson’s “‘reasoning’”— which the Court defines as “the weighing of the [protective] rule’s benefits against its costs,” ante, at 793 — does not justify continued application of the rule it created. The balancing test the Court performs, however, depends entirely on its misunderstanding of Jackson as a rule designed to prevent police badgering, rather than a rule designed to safeguard a defendant’s right to rely on the assistance of counsel.3

*808Next, in order to reach the conclusion that the Jackson rule is unworkable, the Court reframes the relevant inquiry, asking not whether the Jackson rule as applied for the past quarter century has proved easily administrable, but instead whether the Louisiana Supreme Court’s cramped interpretation of that rule is practically workable. The answer to that question, of course, is no. When framed more broadly, however, the evidence is overwhelming that Jackson’s simple, bright-line rule has done more to advance effective law enforcement than to undermine it.

In a supplemental brief submitted by lawyers and judges with extensive experience in law enforcement and prosecution, amici Larry D. Thompson et al. argue persuasively that Jackson’s bright-line rule has provided law enforcement officers with clear guidance, allowed prosecutors to quickly and easily assess whether confessions will be admissible in court, and assisted judges in determining whether a defendant’s Sixth Amendment rights have been violated by police interrogation. See generally Thompson Supplemental Brief 6. While amici acknowledge that “Jackson reduces opportunities to interrogate defendants” and “may require exclusion of evidence that could support a criminal conviction,” they maintain that “it is a rare case where this rule lets a guilty defendant go free.” Ibid. Notably, these representations are not contradicted by the State of Louisiana or other amici, including the United States. See United States Brief 12 (conceding that the Jackson rule has not “resulted in the suppression of significant numbers of statements in federal prosecutions in the past”).4 In short, there is substantial *809evidence suggesting that Jackson’s rule is not only workable, but also desirable from the perspective of law enforcement.

Turning to the reliance interests at stake in the case, the Court rejects the interests of criminal defendants with the flippant observation that any who are knowledgeable enough to rely on Jackson are too savvy to need its protections, and casts aside the reliance interests of law enforcement on the ground that police and prosecutors remain free to employ the Jackson rule if it suits them. See ante, at 793. Again as a result of its mistaken understanding of the purpose behind Jackson’s protective rule, the Court fails to identify the real reliance interest at issue in this case: the public’s interest in knowing that counsel, once secured, may be reasonably relied upon as a medium between the accused and the power of the State. That interest lies at the heart of the Sixth Amendment’s guarantee, and is surely worthy of greater consideration than it is given by today’s decision.

Finally, although the Court acknowledges that “antiquity” is a factor that counsels in favor of retaining precedent, it *810concludes that the fact Jackson is “only two decades old” cuts “in favor of abandoning” the rule it established. Ante, at 792-793. I would have thought that the 23-year existence of a simple bright-line rule would be a factor that cuts in the other direction.

Despite the fact that the rule established in Jackson remains relevant, well grounded in constitutional precedent, and easily administrabie, the Court today rejects it sua sponte. Such a decision can only diminish the public’s confidence in the reliability and fairness of our system of justice.5

Ill

Even if Jackson had never been decided, it would be clear that Monte jo’s Sixth Amendment rights were violated. Today’s decision eliminates the rule that “any waiver of Sixth Amendment rights given in a discussion initiated by police is presumed invalid” once a defendant has invoked his right to counsel. Harvey, 494 U. S., at 349 (citing Jackson, 475 U. S., at 636). Nevertheless, under the undisputed facts of this case, there is no sound basis for concluding that Monte jo made a knowing and valid waiver of his Sixth Amendment right to counsel before acquiescing in police interrogation fol*811lowing his 72-hour hearing. Because police questioned Montejo without notice to, and outside the presence of, his lawyer, the interrogation violated Monte jo’s right to counsel even under pre-Jackson precedent.

Our pre-Jackson case law makes clear that “the Sixth Amendment is violated when the State obtains incriminating statements by knowingly circumventing the accused’s right to have counsel present in a confrontation between the accused and a state agent.” Moulton, 474 U. S., at 176. The Sixth Amendment entitles indicted defendants to have counsel notified of and present during critical confrontations with the State throughout the pretrial process. Given the realities of modern criminal prosecution, the critical proceedings at which counsel’s assistance is required more and more often occur outside the courtroom in pretrial proceedings “where the results might well settle the accused’s fate and reduce the trial itself to a mere formality.” United States v. Wade, 388 U. S. 218, 224 (1967).

In Wade, for instance, we held that because a post-indictment lineup conducted for identification purposes is a critical stage of the criminal proceedings, a defendant and his counsel are constitutionally entitled to notice of the impending lineup. Accordingly, counsel’s presence is a “requisite to conduct of the lineup, absent an intelligent waiver.” Id., at 237 (internal quotation marks omitted). The same reasoning applies to police decisions to interrogate represented defendants. For if the Sixth Amendment entitles an accused to such robust protection during a lineup, surely it entitles him to such protection during a custodial interrogation, when the stakes are as high or higher. Cf. Spano v. New York, 360 U. S. 315, 326 (1959) (Douglas, J., concurring) (“[W]hat use is a defendant’s right to effective counsel at every stage of a criminal case if, while he is held awaiting trial, he can be questioned in the absence of counsel until he confesses?”).

*812The Court avoids confronting the serious Sixth Amendment concerns raised by the police interrogation in this case by assuming that Montejo validly waived his Sixth Amendment rights before submitting to interrogation.6 It does so by summarily concluding that “doctrines ensuring voluntariness of the Fifth Amendment waiver simultaneously ensure the voluntariness of the Sixth Amendment waiver,” ante, at 795; thus, because Montejo was given Miranda warnings prior to interrogation, his waiver was presumptively valid. Ironically, while the Court faults Jackson for blurring the line between this Court’s Fifth and Sixth Amendment jurisprudence, it commits the same error by assuming that the Miranda warnings given in this case, designed purely to safeguard the Fifth Amendment right against self-incrimination, were somehow adequate to protect Montejo’s more robust Sixth Amendment right to counsel.

The majority’s cursory treatment of the waiver question rests entirely on the dubious decision in Patterson, in which we addressed whether, by providing Miranda warnings, police had adequately advised an indicted but unrepresented defendant of his Sixth Amendment right to counsel. The majority held that “[a]s a general matter... an accused who is admonished with the warnings prescribed ... in Miranda, .. . has been sufficiently apprised of the nature of his Sixth Amendment rights, and of the consequences of abandoning those rights.” 487 U. S., at 296. The Court recognized, however, that “because the Sixth Amendment’s protection of the attorney-client relationship . . . extends beyond Mi*813randa’s protection of the Fifth Amendment right to counsel, ... there will be cases where a waiver which would be valid under Miranda will not suffice for Sixth Amendment purposes.” Id., at 297, n. 9. This is such a case.

As I observed in Patterson, the conclusion that Miranda warnings ordinarily provide a sufficient basis for a knowing waiver of the right to counsel rests on the questionable assumption that those warnings make clear to defendants the assistance a lawyer can render during postindictment interrogation. See 487 U. S., at 307 (dissenting opinion). Because Miranda warnings do not hint at the ways in which a lawyer might assist her elient during conversations with the police, I remain convinced that the warnings prescribed in Miranda,7 while sufficient to apprise a defendant of his Fifth Amendment right to remain silent, are inadequate to inform an unrepresented, indicted defendant of his Sixth Amendment right to have a lawyer present at all critical stages of a criminal prosecution. The inadequacy of those warnings is even more obvious in the case of a represented defendant. While it can be argued that informing an indicted but unrepresented defendant of his right to counsel at least alerts him to the fact that he is entitled to obtain something he does not already possess, providing that same warning to a defendant who has already secured counsel is more likely to confound than enlighten.8 By glibly assuming that the Mi*814randa warnings given in this case were sufficient to ensure Montejo’s waiver was both knowing and voluntary, the Court conveniently avoids any comment on the actual advice Montejo received, which did not adequately inform him of his relevant Sixth Amendment rights or alert him to the possible consequences of waiving those rights.

A defendant’s decision to forgo counsel’s assistance and speak openly with police is a momentous one. Given the high stakes of making such a choice and the potential value of counsel’s advice and mediation at that critical stage of the criminal proceedings, it is imperative that a defendant possess “a foil awareness of both the nature of the right being abandoned and the consequences of the decision to abandon it,” Moran v. Burbine, 475 U. S. 412, 421 (1986), before his waiver is deemed valid. See Iowa v. Tovar, 541 U. S. 77, 81 (2004); Johnson v. Zerbst, 304 U. S. 458, 464 (1938). Because the administration of Miranda warnings was insufficient to ensure Montejo understood the Sixth Amendment right he was being asked to surrender, the record in this case provides no basis for concluding that Montejo validly waived his right to counsel, even in the absence of Jackson’s enhanced protections.

IV

The Court’s decision to overrule Jackson is unwarranted. Not only does it rest on a flawed doctrinal premise, but the dubious benefits it hopes to achieve are far outweighed by the damage it does to the rule of law and the integrity of the Sixth Amendment right to counsel. Moreover, even apart *815from the protections afforded by Jackson, the police interrogation in this case violated Jesse Montejo’s Sixth Amendment right to counsel.

I respectfully dissent.

Justice Breyer,

dissenting.

I join Justice Stevens’ dissent except for footnote 5. Although the principles of stare decisis are not inflexible, I believe they bind the Court here. I reached a similar conclusion in Arizona v. Gant, ante, at 354-355 (dissenting opinion), and in several other recent cases. See, e.g., Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U. S. 877, 923-929 (2007) (same); Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701, 865-866 (2007) (same); Federal Election Comm’n v. Wisconsin Right to Life, Inc., 551 U. S. 449, 534-536 (2007) (Souter, J., dissenting); Bowles v. Russell, 551 U. S. 205, 219-220 (2007) (Souter, J., dissenting); Gonzales v. Carhart, 550 U. S. 124, 190-191 (2007) (Ginsburg, J., dissenting); District of Columbia v. Heller, 554 U. S. 570, 675-679 (2008) (Stevens, J., dissenting).