OPINION OP THE COURT

ADAMS, Circuit Judge.

On December 23, 1977, the United States Nuclear Regulatory Commission (NRC) suspended for approximately two years its decisionmaking process regarding proposals for the recycling of spent nuclear fuel and the use in nuclear reactors of plutonium recovered from that fuel. This suspension was announced in an order terminating informal rulemaking and related licensing proceedings concerning this subject. In part, the decision to place a moratorium upon these deliberations was taken in deference to President Carter’s stated objective of deferring domestic plutonium recycling while the United States initiated a multinational evaluation of alternative fuel cycles that would pose a lesser risk of international proliferation of nuclear weapons. Petitions for review were filed requesting us to set aside and enjoin the NRC’s order on the grounds that, in terminating these proceedings, the NRC violated the Atomic Energy Act (AEA)1 and the National Environmental Policy Act (NEPA).2 Because we conclude that the NRC acted within the scope of its authority and that there is no requirement to have a NEPA statement at this time, the petitions for review will be denied.

I.

For over two decades, the federal government, initially through the Atomic Energy Commission (AEC) and later through the NRC, a successor agency to the AEC,3 has been exploring, together with the private sector, the feasibility of reprocessing spent nuclear fuel and employing the plutonium recovered from such fuel in nuclear reactors utilized to generate electricity.4 Commer*763cial implementation of the plutonium recycling process would have the advantages of conserving uranium resources and of alleviating the problem of disposing of radioactive wastes, but might also pose the dangers of a proliferation of nuclear weapons and the possible sabotage of reprocessing facilities. This is so because, unlike the slightly “enriched” uranium currently used in nuclear reactors, plutonium can be employed in the production of nuclear explosives and might be diverted to that end by foreign governments or by terrorists.5

Recognizing that a decision to implement a wide-scale program for the commercial recycling of plutonium constitutes a major federal action significantly affecting the environment, and thereby necessitating an environmental impact statement (EIS) in order to comply with § 102(2)(C) of NEPA,6 the AEC in 1974 commenced work on a Generic Environmental Statement on the Use of Recycle Plutonium in Mixed Oxide Fuel in Light Water Cooled Reactors (GES-MO).7 Concomitant with the GESMO informal rulemaking proceeding, the Commission (this term will be used to include both the AEC and the NRC) conducted adjudicatory licensing proceedings on applications by private companies dealing with the construction and operation of nuclear fuel reprocessing plants, some of which were already pending when GESMO was undertaken. Among the applications before the Commission were those of Allied-General Nuclear Services (Allied-General) for a license to operate the nearly-completed fuel reprocessing plant that it had permission to construct at Barnwell, South Carolina, and of Westinghouse Electric Corp. (Westinghouse) for a license to construct a similar plant at Anderson, South Carolina.

One of the concerns expressed while the rulemaking and adjudicatory proceedings were progressing was that dangers to world security might ensue from the commercial reprocessing of nuclear fuel. For example, the AEC staff’s first draft of GESMO, which was published on August 21, 1974,8 prompted a number of critical comments by the public. These included a letter from the President’s Council on Environmental Quality that was directed at GESMO’s failure (a) to address the proliferation dangers, (b) to explore what safeguards were available, and (c) to weigh the possibility of developing alternative sources of energy.9 In response to this criticism, the staff undertook to reassess its study and to supplement the draft GESMO with an analysis of proliferation risks and safeguards.10 And, on October 28, 1976, President Ford discussed *764the risks entailed in plutonium recycling in a statement on nuclear policy. He declared that the nation “should pursue reprocessing and recycling in the future only if they are found to be consistent with our international [non-proliferation] objectives.” 11

President Carter disclosed his administration’s policy concerning plutonium recycling on April 7, 1977. Noting with alarm the serious proliferation risks of plutonium recycling, the President stated that part of the government’s response would be to “defer indefinitely the commercial reprocessing, and recycling of plutonium produced in the U.S. nuclear power programs,” and to sponsor an international nuclear fuel cycle evaluation (INFCE) program aimed at developing alternative processes with lower proliferation risks.12

Almost immediately thereafter, a motion was filed to terminate the GESMO proceeding, and the NRC’s GESMO Hearing Board postponed further hearings.13 On May 3, 1977, the NRC announced its intention to reassess “the future course and scope of GESMO, the review of recycle-related applications, and the matter of interim licensing,” and invited GESMO participants, the Executive Branch, and other interested persons to submit their views on the subject.14 President Carter’s position was explained on October 4, 1977, in a letter by Stuart Eizenstat, Assistant to the President for Domestic Affairs and Policy.15 The NRC then solicited further public comment on the President’s position and on several alternative courses of action.16

*765Thereafter, on December 23, 1977, the NRC issued an order terminating the GES-MO proceeding as well as most proceedings relating to pending or future plutonium-recycle license applications. Among other things, the order also committed the NRC “to re-examine the above matter after the completion of the ongoing alternative fuel cycle studies, now expected to take about two years,” and to publish shortly after the decision a statement of the reasons underlying the decision.17 That statement, in the form of a Memorandum of Decision, was issued on May 8, 1978.18

Petitioners sought judicial review within the prescribed period of sixty days 19 following entry of the December 23 Order. Westinghouse filed petitions in this Court, which were assigned docket numbers 78-1188 and 78-1189, challenging the termination of both GESMO and the licensing proceedings with respect to its Anderson, South Carolina facility. Four days later, Allied-General (together with Allied Chemical Nuclear Products, Inc., and General Atomic Company) filed a similar petition in the United States Court of Appeals for the district of Columbia objecting to the cessation of GES-MO and licensing proceedings concerning its Barnwell plant. On the same day, Scientists and Engineers for Secure Energy, Mid-Atlantic Legal Foundation, and Capital Legal Foundation (Scientists), filed a petition in this Court, which was assigned docket number 78-1204, seeking review only of the NRC’s termination of GESMO.

Concerned that the December 23 Order might not be deemed final, the parties also took several protective actions after the May 8, 1978, Memorandum of Decision was issued. First, Allied-General petitioned the District of Columbia Court of Appeals for review. Subsequently, Westinghouse and Scientists filed new petitions for relief before this Court, docketed respectively as numbers 78-1895, 78-1894 and 78-1892. Allied-General’s petitions are before us following their transfer from the Court of Appeals of the District of Columbia,20 and have been assigned docket numbers 78-1993 and 78-1994.

To complete the dramatis personae of the immediate controversy, the National Resources Defense Council, Inc. (NRDC) and the States of New York, Texas and Wisconsin have intervened in support of the NRC’s decision, while a group of twenty electric utility companies led by Baltimore Electric and Gas Company have filed an amicus curiae brief urging that the order of the NRC be vacated and remanded for further consideration.

II. CHOICE OF FORUM

NRDC has moved to dismiss for want of jurisdiction those petitions that seek review of the December 23 Order, primarily on the ground that that order is not a reviewable “final order.” 21 At oral argu*766ment it was conceded that granting NRDC’s motion would not affect the entitlement of any party to secure judicial review of the NRC’s decision to terminate the GESMO and related licensing proceedings. This is so because every party that sought review of the December 23 Order also filed a timely petition from the May 8 Memorandum, either in this Court or in the Court of Appeals for the District of Columbia, and if it were determined that the December 23 Order is not “final,” the May 8 Memorandum would perforce constitute such a “final order.” Consequently, no matter which pronouncement is the “final order,” this Court has jurisdiction to review the NRC’s action.22

Apparently, then, what NRDC hopes to achieve by its motions is to have this controversy transferred to the Court of Appeals for the District of Columbia pursuant to 28 U.S.C. § 2112(a). That section specifies which court is the proper forum for review of an administrative decision when petitions are filed in more than one court of appeals. It states in pertinent part:

If proceedings have been instituted in two or more courts of appeals with respect to the same order the agency . shall file the record in thát one of such courts in which a proceeding with respect to such order was first instituted. The other courts in which such proceedings are pending shall thereupon transfer them to the court of appeals in which the record has been filed. For the convenience of the parties in the interest of justice such court may thereafter transfer all the proceedings with respect to such order to any other court of appeals.

Presumably, NRDC believes that if it is concluded that the December 23 Order is not the NRC’s “final order” and that therefore we are not the court “in which a proceeding with respect to such order was first instituted,” we must necessarily transfer all the petitions to the District of Columbia Circuit, which would then decide whether to adjudicate the controversy or perhaps transfer it again to another forum.

Section 2112(a), however, is not jurisdictional in nature, but rather, is a somewhat unusual venue statute.23 It directs that when petitions from “the same order” are filed in different circuits, thereby vesting each of them with jurisdiction, the choice of the appropriate forum for review of that order is to be made by the court in which a petition was filed first. Although it is presumed that that court will retain jurisdiction,24 it remains free to transfer the controversy' elsewhere, “[f]or the convenience of the parties in the interest of justice.” The purpose of § 2112(a) is “to provide a mechanical rule easy of application to avoid confusion and duplication by the courts.” 25 By eliminating the earlier practice whereby the agency selected the forum from among the courts in which petitions had been filed, and by requiring the transfer of all petitions to one court,

“[t]his provision contemplates judicial review of particular agency action by the same court. The parties are spared simultaneous participation in proceedings *767in more than one circuit, and forum shopping is discouraged.26

Although this salutary provision does not by its terms provide a solution for every conceivable situation in which review of the same administrative action is sought in more than one circuit, courts have freely relied on its underlying principles to fashion appropriate approaches to related problems and “[t]o prevent unseemly conflicts that could result should sister circuits each take the initiative and issue conflicting decisions” as to where venue lies.27 Essentially,

[o]ur main concern in handling cases such as this, is that one of the circuits involved take the initiative and decide the threshold question of proper forum rather than delay consideration of the merits due to an overabundance of procedural caution 28

Accordingly, courts have held that even when the validity of the filing of the first petition is in dispute, all the petitions should be transferred to the court in which the disputed petition was filed, and that court should proceed to determine the order in which valid petitions for review were filed.29 Furthermore, when the race to the courthouse results in a virtual “dead heat,” various circuits have adopted the practice of consulting with one another and deferring to one court among them to ascertain, after weighing the interest of justice, which tribunal should be the forum for the decision on the merits.30 The District of Columbia Circuit, which has taken the lead in resolving these procedural problems, has explained its practical approach in the context of a situation where there were virtually simultaneous filings:

In view of the lack of first-filing in any one circuit, Section 2112(a) cannot be implemented through the mechanics set forth in its first sentence quoted above. Instead, we look to the ultimate provision of § 2112(a), which authorizes transfer of the proceedings to another circuit “[f]or the convenience of the parties in the interest of justice.” Judicial review in the forum determined by that standard is a stated congressional goal, and our obligation is- to honor the legislative will as expressed in the statute as a whole.31

Also in keeping with the purposes of the statute, courts have interpreted the term, “the same order,” so as to insure the consolidation in one court of petitions from sequential orders arising from the same administrative background and cumulative record.32

In the present case, the Court of Appeals for the District of Columbia transferred Allied-General’s petitions to this Court notwithstanding the fact that it was not clear whether this Court is the one “in which a proceeding with respect to such order was first instituted.” In granting the NRC’s motion that the petitions be transferred, the District of Columbia Court reasoned in a per curiam opinion that,

[sjince the petitions for review filed in the Third Circuit were filed first, considering the time of filing alone, it seems to us that the (sic) court should decide whether the action of the Commission of December 23 or that of May 8 was its final order in these proceedings. See Abourezk v. FPC, 168 U.S.App.D.C. 246, 513 F.2d 504 (1975)33

*768The District of Columbia Court thus did not base its transfer on the express terms of § 2112(a), but on a judicially-created practice, tailored to meet the statute’s goal of fixing a mechanical rule for determining which court should apply the “interest of justice” standard to choose the forum in which judicial review will take place. Regardless of which court technically had priority under § 2112(a) to choose the appropriate forum for disposition on the merits, our sister court in the District of Columbia graciously ceded to us any right it may have to select that forum so that the possibility of an “unseemly conflict” not arise as to which of the NRC’s actions was its “final order.” Under different circumstances we probably would reach that threshold issue of finality since it is generally expected that the parties will have a fair race to the courthouse.34 In this case, however, we deem it appropriate to retain jurisdiction over these petitions without ultimately deciding which action of the NRC constituted its “final order.”35 We arrive at this conelusion on the basis of our application to the present controversy of the policies that are to be considered in choosing a forum when multiple petitions áre filed for review of agency rulings.

It would do little to further the purposes of § 2112(a) were we now to decide that the May 8 Memorandum was the NRC’s “final order” and therefore transfer all the petitions to the District of Columbia Circuit in order that it may choose a forum. Only if Allied-General, which was the first to file a petition after the May 8 Memorandum (and which filed it in the District of Columbia Court), insisted that the winner of the race to the courthouse be ascertained would it possibly be in the interest of justice to determine the finality question. But in the District of Columbia Circuit, Allied-General did not oppose the NRC’s motion to transfer, and subsequently it expressed its willingness to have this Court adjudicate the validity of the NRC’s action. Allied-General is primarily interested in having the matter resolved expeditiously,36 which certainly *769would not occur were we now to transfer the petitions back to the District of Columbia. This is so because a motions panel of this Court had at an earlier date referred NRDC’s motion to dismiss to a merits panel, and the consolidated cases have already been briefed and argued on the merits.

At oral argument, even counsel for NRDC conceded that the more practical, efficient, and just alternative would be for us to reach the merits. Nonetheless, NRDC did not withdraw its motio.n because it regarded the matter to be a jurisdictional defect that this Court could remedy only if it transferred the cases to the District of Columbia Circuit, which would then presumably retransfer the cases to us for disposition on the merits. We have already concluded that NRDC’s motion does not raise a jurisdictional deficiency because pe-. titions for review from both the December 23 Order and the May 8 Memorandum were timely filed in this Court. There thus being no reason to go through the process of transferring petitions back and forth between two circuits, while the parties as well as the public anxiously await resolution of this important controversy, NRDC’s motion will be denied, and we shall proceed to address the merits.

III. CHALLENGES BASED ON THE ATOMIC ENERGY ACT

A. The Reasons for the NRC Decision

As explained by the NRC in considerable detail in its May 8 Memorandum of Decision, the December 23 Order terminating GESMO and related licensing proceedings was prompted by the President’s policy initiatives as well as by the pendency of studies into alternative fuel cycles. With respect to the first reason for termination, the NRC noted in its Memorandum that although the proceedings in question concern domestic activities, it is appropriate for the NRC to weigh the foreign policy implications as well, because the AEA requires that the common defense and security be considered in making any domestic licensing decision.37 “[I]n the absence of a clear statutory mandate to the contrary” and in view of Congress’ reception of the Administration’s nuclear policy,38 the NRC deemed it proper to accord “substantial deference” to the President’s request that it terminate the proceedings, since “the President is the national spokesman in the area of foreign policy.”39 In addition, the NRC examined the President’s request and reasoned that indeed the country’s international prerogatives would be compromised were the proceedings to continue.40

*770The second reason given for termination at this time is the pendency of two alternative fuel cycle studies — INFCE, which is expected to take at least two years, and an interagency Non-Proliferation Alternative System Assessment Program (NASAP) under the coordination of the Department of Energy, which is being undertaken to develop a technical basis for the United States’ contributions to INFCE. The NRC concluded that because the results of these studies are anticipated in the near term, and because the NRC must consider the information and alternatives developed therein before reaching a decision on the pending applications or on GESMO, it would be a waste of effort to continue the pending proceedings and then to supplement or completely revise the record to reflect those results.41

The NRC carefully examined policy-based arguments tendered by various commentors that the proceedings ought to be continued, but was not persuaded by them.42 It emphasized that its decision to terminate the GESMO and related licensing proceedings “does not involve their final disposition on the merits,” and is taken because “the present state of studies and national fuel cycle policy evaluations precludes an informed decision on the merits of plutonium recycle at this time.”43 Finally, the NRC noted that the President and Congress have indicated that they may reassess their positions after the studies are completed, and that the NRC is committed to reexamine its December 23 decision “in light of the completed studies, expected to take about two years, and any revisions of the Administration’s policies.” 44

Essentially, then, the NRC has imposed a moratorium, expected to last about two years, upon its decisionmaking process regarding plutonium recycling, and, as a result, has terminated the pending GESMO and related licensing proceedings. In the agency’s view, it had “the discretion to stop processing applications and to refuse to accept new ones when there are sound regulatory reasons to do so.”45 We shall assess the validity of the NRC’s assertion by examining first, whether the imposition of a moratorium contravenes the legislative scheme, and second, whether the NRC abused its discretion, abridged procedural requirements, or acted arbitrarily, capriciously, or otherwise not in accordance with the law when it issued the December 23 Order.46

*771B. The NRC’s Authority to Place a Moratorium Upon a Decisionmaking Process Conducted Pursuant to the AEA and thereby to Terminate Rulemaking and Licensing Proceedings

In passing the Atomic Energy Act of 1954, Congress enacted “a regulatory scheme which is virtually unique in the degree to which broad responsibility is reposed in the administering agency, free of close prescription in its charter as to how it shall proceed in achieving the statutory objectives.” 47 This legislation declares it to be the policy of the United States that the development, use, and control of atomic energy be conducted in such a way as “to promote world peace” and “to make the maximum contribution to the general welfare, subject at all times to the paramount objective of making the maximum contribution to the common defense and security.”48 The AEA is designed to effectuate this policy by providing for a wide range of programs. Among the programs specified in § 3, and particularly relevant to the present case, are:

(c) a program for Government control of the possession, use, and production of atomic energy and special nuclear material, whether owned by the Government or others, so directed as to ... . provide continued assurance of the Government’s ability to enter into and enforce agreements with nations or groups of nations for the control of special nuclear materials and atomic weapons; . [and]

(e) a program of international cooperation to promote the common defense and security . . . ,49

To further these statutory objectives, § 161(p) of the AEA50 confers upon the NRC an unfettered mandate to issue “such rules and regulations as may be necessary to carry out the purposes of this Act.” With slightly greater specificity, and relevant for the GESMO' rulemaking proceeding, § 161(b) authorizes the Commission to

establish by rule, regulation, or order, such standards and instructions to govern the possession and use of special nuclear material, source material, and byproduct material as the Commission may deem necessary or desirable to promote the common defense and security or to protect health or to minimize danger to life or property . . . .51

A third provision, § 103,52 directs that commercial licenses, such as those for which applications have been made in the present case, shall be issued “subject to such conditions as the Commission may by rule or regulation establish to effectuate the purposes and provisions of this chapter.” Section 103 admonishes, however, that “no license may be issued . . if,y in the opinion of the Commission, the issuánce of a license to such person would be inimical to the common defense and security or to the health and safety of the public.”

Given this broad delegation of authority to the NRC to choose the necessary means by which to implement the general policy objectives of the AEA, we cannot say that the NRC must inexorably proceed with the processing of license applications and the development of a final GESMO when in its judgment to do so would endanger the attainment of its statutory objectives. Indeed, it would appear to be fully congruent with the broad delegation of authority in the AEA, which allows the NRC to determine the conditions, rules and regulations pursuant to which licenses shall be issued as well as the scope and format of licensing *772proceedings, to construe that statute’s provisions as granting the NRC the authority to refuse to conduct a licensing proceeding for a reasonable period of time in furtherance of its statutorily delineated responsibilities.

Our conclusion that, an agency that is invested with such extensive powers to effectuate its far-reaching mandate may impose a moratorium upon its decisionmaking process when sound regulatory reasons exist for doing so is, we believe, consonant with the Supreme Court’s approach in Permian Basin Area Rate Cases, 390 U.S. 747, 88 S.Ct. 1344, 20 L.Ed.2d 312 (1968). There, the question was whether the Federal Power Commission (FPC) had authority to impose a two and one-half year moratorium upon filings of rate schedules while it implemented a new regional ratemaking scheme. The Court noted that it “has repeatedly held that the width of administrative authority must be measured in part by the purposes for which it was conferred . .”53 Citing § 16 of the Natural Gas Act,54 which is similar in breadth to § 161(p) of the AEA, the Court continued, “[s]urely the Commission’s broad responsibilities therefore demand a generous construction of its statutory authority.” 55 After rejecting arguments for a narrow construction of the FPC’s authority that were based on a number of statutory provisions, the Court concluded that a moratorium may be imposed. It stated that, “[w]e are, in the absence of compelling evidence that such was Congress’ intention, unwilling to prohibit administrative action imperative for the achievement of an agency’s ultimate purposes.”56

Nothing brought to our attention by petitioners regarding the AEA or its history has convinced us that the NRC has less discretion to impose a moratorium in appropriate circumstances than does the FPC, or that the NRC has failed to comply with any procedural requirements in deferring consideration of wide-scale plutonium recycling. We agree with petitioners that under § 103, once an applicant complies with the provisions of the AEA and Commission rules and regulations, the NRC must issue a license unless it determines that “the issuance of a license to such person would be inimical to the common defense and security or to the health and safety of the public.” 57 But we do not believe that a finding of inimicality or noncompliance with the applicable requirements has to be made before the NRC may suspend license application proceedings. This would appear to be particularly true where a moratorium is declared to- enable the Commission to make a reasoned decision regarding the rules and regulations that should be applied and whether the issuance of licenses would be inimical to the common defense and security-

Nor are we persuaded by petitioners’ contention that in suspending the license application proceedings the NRC violated § 189(a) of the AEA,58 which requires that a public hearing be afforded in any license proceeding “upon the request of any person whose interest may be affected by the proceeding.” As we understand it, that provision is similar to one governing FCC licensing procedures.59 It guarantees a hearing *773to all interested parties sometime before the NRC decides whether to grant or deny a license application, it does not mandate that the proceedings in which the ultimate disposition of such an application is being considered continue without any interruption or that a hearing be held before the NRC orders that a moratorium be placed upon its decisionmaking process.

C. The NRC’s Exercise of its Discretion to Impose a Moratorium Upon its Decisionmaking Process and to Terminate GESMO and Related Licensing Proceedings

Inasmuch as we conclude that the NRC may, in its discretion, delay processing applications and refuse to accept new ones when there are sound reasons for doing so, we must now examine whether in the present case this discretion was abused or was exercised in a manner not in conformity with Statutory dictates. Petitioners raise four grounds upon which the NRC’s decision may be faulted.

m First, petitioners suggest that under the due process clause they should have been afforded an adjudicatory-type hearing before the NRC issued the December 23 Order. However, cases upholding the imposition of temporary “freezes” upon the processing of license applications by the FCC 60 indicate that applicants need not be afforded such a hearing before an agency suspends the processing of license applications. Here, the NRC twice gave notice and solicited comments from all affected parties pri- or to releasing its December 23 Order. In the absence of any other statutory or regulatory prescriptions, it is not for the courts to impose additional procedural requirements on the agency.61

Next, petitioners assert that in fact the NRC did not place a moratorium upon its decisionmaking process regarding plutonium recycling but really terminated it forever and thus effectively denied the license applications, subject only to the possibility of future reconsideration. In support of this interpretation of the NRC decision, petitioners point out that despite Commissioner Kennedy’s objections,62 the NRC decided to “terminate,” rather than to “defer” the GESMO and related licensing proceedings. The NRC, it is further noted, did not guar*774antee that it would reopen the proceedings in two years, and has conceded in its Memorandum that “it is not possible to determine whether our proceedings will then be reinstituted or whether some other course will be adopted.” 63

This contention fails to distinguish between the GESMO and related licensing proceedings, on the one hand, and the ultimate issues of whether to grant the licenses and whether to commit the United States to plutonium recycling programs on the other hand. The December 23 decision terminated the GESMO and related licensing proceedings because in the judgment of the NRC the ultimate issues could not be resolved on their merits within the next two years. At the same time, the NRC committed itself to reexamine this decision in light of future developments, and to determine at a later date what course of action should then be taken on the license applications and the overall question of wide-spread plutonium recycling. Presumably, whatever course the NRC does adopt in the future will lead to a final determination whether or not to issue the requested licenses. In view of the flexibility that the NRC has in fixing the scope and format of its proceedings, as well as the real possibility that the GESMO and licensing proceedings as originally scheduled will be an inappropriate vehicle for arriving at a decision on the ultimate issues, we cannot say that the NRC abused its discretion in “terminating” rather than “deferring” the present proceedings and in refusing to bind itself to a future plan. Furthermore, as we understand it, such a decision may properly be regarded as a statutorily permitted moratorium on the decisionmaking process regarding the ultimate issues, which are still before the agency. Finally, under the circumstances, the expected two-year duration of this moratorium cannot be deemed unreasonable.

Of course, the NRC may not completely terminate license application proceedings without passing on the merits of the applications, simply by declaring an open-ended moratorium. It is required by statute to fix the conditions and regulations pursuant to which licenses will be granted, and to award such licenses if the prerequisites are met, unless it makes a finding of inimicality to the common defense and security or to the public health and safety.64 But we are satisfied, at least for now, that the NRC has not abused its discretion in refusing to continue the pending proceedings on the ground that it cannot yet formulate a generic standard or make a determination on the question of inimicality. When and if it ever becomes apparent that the NRC has de facto denied the license applications despite the applicants’ compliance with the pertinent regulations and without making a finding of inimicality, or that the moratorium is of unreasonable duration, judicial recourse will be available to the aggrieved parties.

Petitioners also maintain that the NRC impermissibly terminated the GESMO and related licensing proceedings at the request of the President and in deference to his foreign policy pronouncement. They charge that in failing to act independently of the Executive Branch, the NRC contravened Congress’ express intent that the Commission be completely free from presidential influence and control.65 According *775to petitioners, the fact that the President has primary responsibility over foreign affairs does not justify the NRC’s derogation from Congress’ plan, since foreign affairs powers are vested in Congress as well. And, petitioners assert, Congress exercised its share of those powers by legislating that with respect to nuclear energy, which inevitably touches upon the sensitive area of foreign affairs, a strict separation is to prevail between the President and the Commission. This is particularly so, petitioners declare, in the context of domestic licensing. In contrast, in the area of international arrangements, such as export licensing, Congress has legislated a divergence from this scheme by making Commission decisions subject to approval by the President.66

Although petitioners’ argument is resonant with constitutional subtleties concerning the “twilight zone” 67 in which congressional and presidential powers overlap, neither the authority reposed in the NRC nor the agency’s exercise of that authority in this case require us to venture into that largely unchartered area. The legislative history produced by petitioners makes clear that Congress intended that the Commission be independent not only from pressures brought to bear by the President, but from all external pressures. Representative is the view expressed by Senator Magnuson:

Actually, the AEC was established by Congress with the hope and aim of making it the most sensitive agency of Government, more independent than any other, and to be protected from Congress itself, and from all other interference, including Executive interference.68

Independence, however, does not mean that the Commission must ignore or reject positions espoused by the President, by Congress or by other parties. The Commission was “charged with a most sensitive and most vital responsibility,” 69 a responsibility that cannot possibly be performed properly if the Commission is oblivious or nonresponsive to actions being taken by others, whether within or outside the government. When it created independent administrative agencies, Congress undoubtedly desired that they interact with the three branches of the government much as the legislative body interacts with the executive branch, with “separateness but interdependence, autonomy but reciprocity,” so that “practice will integrate the dispersed powers into a workable government.”70

As previously set forth, the NRC is directed in many provisions of the AEA to consider “the common defense and security.” Any contemplation of these sensitive matters necessarily touches upon areas that are also within the domain of the President and of Congress.71 It was therefore appropriate for the NRC to take note of the relevant developments in the executive and legislative branches and to ascertain, with the help of interested parties, what bearing these developments may have on its own *776agenda. As we understand the NRC’s actions here, that is all it did, and it maintained its independence from both those branches while making an informed decision to suspend its proceedings.

There is no evidence that the President improperly interfered with the NRC’s decisionmaking process, or that the NRC capitulated to the President. Instead, the agency appears to have examined the President’s position, and agreed with the President’s contention that continuation of the proceedings would adversely affect the President’s nonproliferation efforts. Then, after determining that Congress had not exercised its constitutional powers in this area in a contrary manner, neither through the AEA nor through subsequent legislation, the NRC decided that it would be prudent to terminate the proceedings for a time so that the President might pursue his objectives. Given this record, we cannot say that the NRC abused its discretion or acted arbitrarily, capriciously, or not in accordance with the law when it rested its decision in part on a desire not to obstruct the goal of securing international nonproliferation.

As a final ground for attacking the NRC’s exercise of its discretion in the present case, petitioners contend that the NRC impermissibly relied upon the INFCE and NASAP studies in deciding to terminate GESMO and the related licensing proceedings. This is so, they argue, because assertions about these studies are made in the Memorandum without support in the record, in violation of 5 U.S.C. § 556(e),72 and also because the studies are not being undertaken by the NRC and are beyond its control. Petitioners would have us confine an agency’s discretion to prescribe a moratorium upon its decisionmaking process while the results of relevant studies are awaited, to those situations where the agency itself is conducting the investigation, as occurred in the Permian Basin Rate Cases and the FCC “freeze” cases. But to place such a limitation upon the NRC’s discretion to impose a moratorium would appear to serve no regulatory purpose. If the NRC may freeze pending proceedings when in its judgment it cannot make a reasoned decision until the results of ongoing studies are completed and available for evaluation, it should make no difference whether the studies are being prepared by the NRC or by other parties. As for the particular studies involved, they were identified and mentioned throughout the proceedings leading up to the December 23 Order, and the NRC did not violate 5 U.S.C. § 556(e) in relying on them as a partial basis for its decision.

IV. CHALLENGES BASED ON THE NATIONAL ENVIRONMENTAL POLICY ACT

Aside from attacking the NRC’s decision of December 23 on the ground that it violated the AEA, petitioners charge that the NRC was required under § 102(2)(C) of NEPA73 to prepare an environmental impact statement (EIS) before deciding not to proceed with wide-scale plutonium recycle and to terminate the GESMO and related licensing proceedings, since that decision constituted a major federal action significantly affecting the quality of the human environment. For an appreciation of the scope of § 102(2)(C), petitioners refer us to Scientists’ Institute for Public Information, Inc. v. AEC, 156 U.S.App.D.C. 395, 481 F.2d 1079 (1973), where the court stated:

The statutory phrase “actions significantly affecting the quality of the environment” is intentionally broad, reflecting the Act’s attempt to promote an across-the-board adjustment in federal agency decision making so as to make the quality *777of the environment a concern of every federal agency. The legislative history of the Act indicates that the term “actions” refers not only to construction of particular facilities, but includes “project proposals, proposals for new legislation, regulations, policy statements, or expansion or revision of ongoing programs ” 74

Petitioners also cite National Helium Corp. v. Morton, 455 F.2d 650 (10 Cir. 1971), for the proposition that it makes no difference for purposes of NEPA whether the decision is to terminate rather than to institute a program. And, petitioners point out, it is well settled that a post hoc rationalization will not satisfy NEPA’s procedural mandate.75

Turning to the December 23 Order, petitioners insist that inasmuch as any decision whether or not to authorize wide-scale plutonium recycle goes to the heart of the substantive policy of § 101(b)(6) of NEPA76 (namely, that the recycling of depletable resources be maximized), compliance with the procedural requirement that the NRC prepare an EIS is particularly important.77 Petitioners point out that the NRC itself has on many occasions recognized the need to compose an EIS on this subject, and that it undertook to prepare GESMO to fulfill that statutory obligation.78 Inasmuch as any decision on either the wide-scale deployment of plutonium recycling facilities or any of the individual licensing applications would significantly affect the environment, petitioners contend, the NRC was required to complete GESMO prior to any decision not to authorize plutonium recycling. Further, petitioners assert that the December 23 Order was a decision not to authorize wide-scale plutonium recycling because it brought to a halt all progress toward the implementation of this new technology after significant prior federal involvement, and denied the applicants any means of pursuing their license applications.

It is already apparent from our analysis of petitioners’ challenges under the AEA that we view the December 23 Order suspending the decisionmaking process and terminating GESMO and related licensing proceedings in a different light. Contrary to petitioners’ contentions, we have already concluded that the NRC has not yet decided against authorizing wide-scale plutonium recycling and to deny the license applications. Rather, it has ordered a cessation, for a period expected to last approximately two years, of its decisionmaking process on these issues. We believe it evident that the NRC was not required to prepare an EIS before taking this step, for a number of reasons:

The December 23 Order must be seen in the context of the entire course of events contemplated by the NRC. After it received a number of applications for licensing of plutonium recycling facilities, the NRC realized that it was faced with policy choices that would have a significant impact on the environment, and that it was therefore necessary to prepare an EIS before making any decision whether to authorize the wide-scale reprocessing of spent fuel or to grant the requested licenses. GESMO was to be that EIS, and was to be completed prior' to any decision so that it could introduce into the NRC’s decision-making process input on relevant environmental issues, as mandated by NEPA.79 What the December 23 Order did was to *778suspend the decisionmaking process, including the preparation of an EIS, without taking any position on the ultimate issues for which an EIS is required. This action in no way violates NEPA. As the Supreme Court declared in Kleppe v. Sierra Club, 427 U.S. 390, 96 S.Ct. 2718, 49 L.Ed.2d 576 (1976), “[T]he procedural duty imposed upon agencies by [§ 102(2)(c)] is quite precise, and the role of the courts in enforcing that duty is similarly precise.”80 An EIS is not required, the Supreme Court continued, when an agency is “merely contemplating] . [a proposal or] at any other point prior to the formal recommendation or report on a proposal,” because “[c]ontemplation of a project and the accompanying study thereof do not necessarily result in a proposal for major federal actions.”81 Moreover, the Court counseled that it is not the role of the judiciary to “determine a point during the germination process of a potential proposal at which an impact statement should be prepared.” 82 There is nothing in the record to suggest that the NRC will not undertake to complete an EIS as part of its decisionmaking process on whether or not to authorize the wide-scale recycling of plutonium and whether or not to grant the requested licenses. The specific time when that EIS is to be prepared, however, is for the NRC to determine.83

It is also important to note that petitioners’ argument in this respect would produce a completely illogical result. As just mentioned, GESMO itself was to be an EIS. The December 23 Order, which terminated the GESMO proceedings, was based in part on the NRC’s judgment that it could not make an informed decision — as it is required to do under NEPA 84—on the environmental impact of reprocessing spent fuel because studies are still in progress on the feasibility of employing alternative fuel sources. Not only would the adoption of petitioners’ position necessitate the preparation of an EIS — the statutorily-compelled procedure for ensuring informed decision-making on matters affecting the environment — despite this lack of information, it would also require the preparation of an EIS before an agency may delay preparation of an EIS. Such a result could not have been intended by Congress when it enacted NEPA.

V. CONCLUSION

A pressing problem in this last quarter of the twentieth century is a dwindling supply of energy resources coincident with a rising demand for that supply. Increased reliance has been placed on nuclear fuel as one of the more promising sources of energy, at least for the next few years. Nuclear energy, however, presents problems of its own. As recent events have demonstrated, radiation danger to the general population cannot always be contained. Moreover, there exists the risk that nuclear materials and knowledge will be used for nonpeaceful purposes.

Far-reaching policy choices must inevitably be made regarding energy resources, *779development, conservation and consumption. Indeed, the directions selected may well determine the future character of our society. In a democracy, though, these choices must be made by the political branches of government, not by the courts.

Congress has delegated authority in the delicate area of nuclear energy to a number of agencies, among them the NRC. The NRC is charged with the responsibility of protecting the common defense and security as well as the public health and safety, while overseeing the licensing of nuclear facilities. Some of the decisions it makes to further its statutory mandate may be unpopular in the nuclear industry, among environmentalists, or with other groups of citizens. But Congress has decreed that the agency be independent from outside control, and it would subvert this design were we to invalidate the challenged NRC action when it appears to be consonant with statutory dictates and not an unreasonable exercise of its discretion.

Accordingly, the petitions for review will be denied.

*398Mr. Justice Brennan

delivered the opinion of the Court.

This case is the first contested licensing proceeding to be decided by the Atomic Energy Commission under the Atomic Energy Act of 1954, 68 Stat. 919, 42 U. S. C. § 2011 et seq. It presents the question whether the Commission erred in continuing in effect a provisional construction permit which authorizes the petitioner Power Reactor Development Company to construct, but not to operate, a fast-neutron breeder reactor for the generation of electric power. The Court of Appeals for the District of Columbia Circuit set that order aside. 108 U. S. App. D. C. 97, 280 F. 2d 645 (1960). We granted certiorari, 364 U. S. 889 (1960), on petitions of the United States and of Power Reactor Development Company (hereafter PRDC), to decide an important question of the scope of the Commission’s power under the Atomic Energy Act of 1954.

Stated more precisely, the question before us is whether the Commission, in issuing a permit for the construction of a facility which will utilize nuclear materials, such as the power reactor presently involved, must make the same definitive finding of safety of operation as it admittedly will have to make before it licenses actual operation of the facility. The Court of Appeals said: “It is undisputed that the Commission must make such a finding when it authorizes operation. The question is whether it must make such a finding when it authorizes construction. In our opinion it must.” 108 U. S. App. D. C., at 100, 280 F. 2d, at 648. Petitioners agree that some finding directed to safety of operation must be made at the construction-permit stage of the proceeding, but argue that the Court of Appeals erred in holding that the Commission must have the same degree of certitude at this preliminary point as when it licenses operation. In order to understand how the controversy arises and what is involved in *399its resolution, it will be necessary to state the proceedings in the case at some length, and then describe in detail the governing statute and administrative regulations. For the decision of this case ultimately turns on a comparison of what the Commission found with what the statute and regulations require.

The case began on January 7, 1956, when PRDC filed with the Commission (hereafter sometimes referred to as the AEC) an application to construct and operate a developmental power reactor of a relatively new type. This device has two characteristics which distinguish it from other nuclear reactors. First, the neutrons which fly about inside the reactor (to use crude but graphic layman’s terminology) and split atoms of fissionable Uranium-235 — thus releasing new neutrons and energy in the form of heat — are “fast” neutrons. That is, they travel at a velocity of about 10,000 miles per second, much faster than neutrons in ordinary reactors. Second, this réactor is a “breeder”: it has the property of being able to produce about 1.2 times as much fissionable material as it consumes. This result comes about through a sort of modern alchemy; when the neutrons fly outside the inner core of the reactor, which is composed of fissionable U-235, they enter a blanket of nonfissionable U-238. Atoms in this blanket are changed, when struck by a neutron, into Plutonium, itself a fissionable fuel which can be removed from the reactor and be put to possible use in other installations. Thus, the reactor “breeds” Plutonium faster than it uses up U-235. It not only generates energy to produce electric power, it also creates new reactor fuel. This “breeder” effect is attainable because of the use of fast neutrons. Two boron control rods inserted into the reactor are a means designed to reduce its power level at any time. And in addition to these rods, eight more boron rods are suspended by an electromagnet over the reactor; in case the reactivity rises to a danger*400ously high level, these safety rods are intended to drop into the reactor automatically and shut it down immediately. The whole machine is housed in a series of thick concrete, graphite, and steel layers, all underground. Over this entire complex is placed a football-shaped building, enclosed in a two-inch steel shield capable of containing an explosion equal in force to 1,000 pounds of TNT, which is greater than any explosion which any of the experts who testified in this case believes is at all likely to result from an accident in the operation of the reactor. The application, after describing the reactor in much greater detail than this rudimentary summary, went on to provide that the reactor would be located at Lagoona Beach, Mich., on the shores of Lake Erie, about 35 miles from the center of Detroit, Mich., and about 30 miles from the center of Toledo, Ohio.

The Commission took the case under advisement and, on August 4, 1956, despite a report of its Advisory Committee on Reactor Safeguards which was at best noncommittal about the probable safety of the proposed reactor in operation, issued a provisional construction permit without having held public hearings, as the law at that time permitted it to do. This permit was subject to the following condition:

“The conversion of this permit to a license is subject to submittal by PRDC to the Commission (by amendment of the application) of the complete, final Hazards Summary Report (portions of which may be submitted and evaluated from time to time). The final Hazards Summary Report must show that the final design provides Reasonable assurance . . . that the health and safety''of the public will not be endangered by operation of the reactor

On August 31, 1956, in accordance with the Commission’s then existing rules of practice, the respondents in *401this Court, International Union of Electrical, Radio, and Machine Workers, United Automobile, Aircraft, and Agricultural Implement Workers of America, and United Papermakers and Paperworkers, petitioned the Commission for permission to intervene and oppose continuation in effect of PRDC’s provisional construction permit. The AEC granted permission to intervene on October 8, 1956, and set the case down for a hearing before one of its hearing examiners. Extensive hearings were held between January 8, 1957, and August 7, 1957, and on November 22, 1957, in accordance with the AEC’s order setting the case for hearing before him, the examiner, instead of issuing an initial decision and opinion of his own, transferred and certified the record of the hearings to the full Commission for its consideration. Oral argument was had before the Commission on May 29, 1958. On December 10, 1958, the Commission rendered its “Opinion and Initial Decision” continuing PRDC’s permit in effect, subject to the same condition recited above. To its opinion were appended extensive findings of fact, including Finding 22, which is of central importance to the decision of this case. That finding reads as follows:

“22. The Commission finds reasonable assurance in the record that a utilization facility of the general type proposed in the PRDC application and amendments thereto can be constructed and will be able to be operated at the location proposed without undue risk to the health and safety of the public.”

Commissioners Vance and Floberg joined in the opinion. Commissioner Graham filed a short concurring opinion agreeing with the Commission’s basic safety findings, just quoted, but doing so in much shorter compass than the majority. Commissioners Libby and McCone (the chairman) took no part in the decision. The result of this initial opinion was an order continuing PRDC’s provi*402sional construction permit in effect, but containing the same condition which the original permit, issued on August 4, 1956, had contained.

The intervening unions, as was their right, filed detailed exceptions to this initial decision. The Commission fully reconsidered all the contentions and reviewed the evidence presented at the lengthy hearings, with particular attention to the testimony of the scientific experts, several of them members of the Advisory Committee on Reactor Safeguards, who had testified. On May 26,1959, the Commission issued its “Opinion and Final Decision,” dealing with all questions presented in even greater detail and reaffirming its initial decision. The Commission emphasized that “public safety is the first, last, and a permanent consideration in any decision on the issuance of a construction permit or a license to operate a nuclear facility.” Even after operation of the reactor is licensed — if it ever is — the Commission, it said, will retain jurisdiction over PRDC’s activities to ensure that the highest safety standards are maintained. The opinion went on to examine the suitability of the proposed site, noted that it was near a great population center, and nevertheless concluded that at the present stage there was reasonable assurance that the general type of reactor proposed by PRDC would be safe enough at that location. The Commission pointed out, however, that its action in allowing PRDC to proceed with construction was by its nature tentative and preliminary, and that it was by no means committed to the issuance of an operating license. “PRDC has been on notice since before the first shovel of dirt was moved,” it said, “that its construction permit is provisional upon further demonstration of many technological and financial facts, including the complete safety of the reactor.” A more severe safety test would have to be passed when the reactor was completed, the opinion said, since “[t]he degree of ‘reasonable assurance’. . . *403that satisfies us . . . for purposes of the provisional construction permit would not be the same as we would require in considering the issuance of the operating license.” The Commission then made new findings of fact, including the following counterpart of its initial Finding 22:

“22. The Commission finds reasonable assurance in the record, for the purposes of this provisional construction permit, that a utilization facility of the general type proposed in the PRDC Application and amendments thereto can be constructed and operated at the location without undue risk to the health and safety of the public.”

All three of the Commissioners who took part in the case joined in this final decision, and the Commission entered its final order continuing in effect the PRDC provisional construction permit, but again subject to the condition that a more extensive safety investigation, and a definitive safety finding, would have to be made before operation was permitted.

The intervening unions, respondents in this Court, then petitioned the Court of Appeals for the District of Columbia Circuit to review and set aside this order of the Commission. Only the final order continuing the permit in effect was drawn in question. No complaint was made of the original ex parte grant of the permit in 1956. PRDC intervened in the Court of Appeals in support of the AEC. On June 10, 1960, by a divided vote, a three-judge panel of the Court of Appeals set aside the AEC’s order and remanded the case to the Commission. A petition for rehearing en banc was denied, two judges dissenting, and we brought the case here.

We turn now to an examination of the statutes and regulations pursuant to which the Commission purported to continue in effect PRDC’s construction permit. The *404basic provision is § 104b of the Atomic Energy Act of 1954, 42 U. S. C. § 2134 (b), which authorizes the AEC to “issue licenses to persons applying therefor for utilization and production facilities involved in the conduct of research and development activities .... In issuing licenses under this subsection, the Commission shall impose the minimum amount of such regulations and terms of license as will permit the Commission to fulfill its obligations under this chapter to promote the common defense and security and to protect the health and safety of the public . . . .” Two things about this section should be emphasized. First, there is no doubt that the term “licenses” as used therein includes the provisional construction permit which PRDC has received. The last sentence of § 185, 42 U. S. C. § 2235, expressly so provides, as we shall soon see. And second, there is also no doubt that construction permits, like all other licenses, can be issued only consistently with the health and safety of the public. But the responsibility for safeguarding that health and safety belongs under the statute to the Commission. And § 104b, especially when read in connection with the general rule-making power conferred by § 161i (3), 42 U. S. C. § 2201 (i) (3), clearly contemplates that the Commission shall by regulation set forth what the public safety requires as a prerequisite to the issuance of any license or permit under the Act.

The issuance of construction permits is subject to § 185, 42 U. S. C. § 2235. That section provides that

“All applicants for licenses to construct or modify production or utilization facilities shall, if the application is otherwise acceptable to the Commission, be initially granted a construction permit. The construction permit shall state the earliest and latest dates for the completion of the construction or modification. Unless the construction or modification of the facility is completed by the completion date, the *405construction permit shall expire, and all rights thereunder be forfeited, unless upon good cause shown, the Commission extends the completion date. Upon the completion of the construction or modification of the facility, upon the filing of any additional information needed to bring the original application up to date, and upon finding that the facility authorized has been constructed and will operate in conformity with the application as amended and in conformity with the provisions of this chapter and of the rules and regulations of the Commisson, and in the absence of any good cause being shown to the Commission why the granting of a license would not be in accordance with the provisions of this chapter, the Commission shall thereupon issue a license to the applicant. For all other purposes of this chapter, a construction permit is deemed to be a ‘license.’ ”

It is clear from the face of this statute — and all parties agree — that Congress contemplated a step-by-step procedure. First an applicant would have to get a construction permit, then he would have to construct his facility, and then he would have to ask the Commission to grant him a license to operate the facility. This procedure is described in its general outlines in Marks and Trowbridge, Framework for Atomic Industry, 76-77 (1955). See also Green, The Law of Reactor Safety, 12 Yand. L. Rev. 112, 121-127 (1958). The second step of the procedure, the application for and granting of an operating license, is governed by § 182a, 42 U. S. C. § 2232 (a). That provision reads, in pertinent part:

“In connection with applications for licenses to operate production or utilization facilities, the applicant shall state such technical specifications . . . and such other information as the Commission may, by rule or regulation, deem necessary in order to enable it to find that the utilization or production of special *406nuclear material will be in accord with the common defense and security and will provide adequate protection to the health and safety of the public.”

It is clear from this provision that before licensing the operation of PRDC’s reactor, the AEC will have to make a positive finding that operation of the facility will “provide adequate protection to the health and safety of the public.” What is not clear, and what is at the center of the controversy in this case, is whether the Commission must also have made such a finding when it issued PRDC’s construction permit. There is nothing on the face of either § 182 or f 185 which tells us what safety findings must be made before this preliminary step is taken. We know, however, from § 104b that some such finding must be made. For enlightenment on the nature of this finding, both parties urge us to examine the Commission’s regulations, and accordingly we proceed to do so.

The crucial regulation for our purposes is the Commission’s regulation 50.35, 10 CFR § 50.35:

“§ 50.35. Extended time for providing technical information. Where, because of the nature of a proposed project, an applicant is not in a position to supply initially all of the technical information otherwise required to complete the application, he shall indicate the reason, the items or kinds of information omitted, and the approximate times when such data will be produced. If the Commission is satisfied that it has information sufficient to provide reasonable assurance that a facility of the general type proposed can be constructed and operated at the proposed location without undue risk to the health and safety of the public and that the omitted information will be supplied, it may process the application and issue a construction permit on a provisional basis without the omitted information subject *407to its later production and an evaluation by the Commission that the final design provides reasonable assurance that the health and safety of the public will not be endangered.”

This regulation, obviously, elaborates upon and describes in fuller detail the step-by-step licensing procedure contemplated by §§ 182 and 185. It states, pursuant to the authority conferred by §§ 104b and 161i (3), what safety findings shall be required at each stage of the proceeding. There is general agreement that the second safety finding referred to, “that the final design provides reasonable assurance that the health and safety of the public will not be endangered,” comports with the requirements of § 182 concerning the issuance of a license to operate. There is also agreement that the regulation’s first required safety finding, “that [the AEC] has information sufficient to provide reasonable assurance that a facility of the general type proposed can be constructed and operated at the proposed location without undue risk to the health and safety of the public,” is a valid exercise of the rule-making power conferred upon the AEC by statute, and requires that some finding as to safety of operation be made even before a provisional construction permit is granted. The question is whether that first finding must be backed up with as much conviction as to the safety of the final design of the specific reactor in operation as the second, final finding must be.

We think the great weight of the argument supports the position taken by PRDC and by the Commission, that Reg. 50.35 permits the Commission to defer a definitive safety finding until operation is actually licensed. The words of the regulation themselves certainly lean strongly in that direction. The first finding is to be made, by definition, on the basis of incomplete information, and concerns only the “general type” of reactor proposed. *408The second finding is phrased unequivocally in terms of “reasonable assurance,” while the first speaks more tentatively of “information sufficient to provide reasonable assurance.” The Commission, furthermore, had good reason to make this distinction. For nuclear reactors are fast-developing and fast-changing. What is up to date now may not, probably will not, be as acceptable tomorrow. Problems which seem insuperable now may be solved tomorrow, perhaps in the very process of construction itself. We see no reason why we should not accord to the Commission's interpretation of its own regulation and governing statute that respect which is customarily given to a practical administrative construction of a disputed provision. Particularly is this respect due when the administrative practice at stake “involves a contemporaneous construction of a statute by the men charged with the responsibility of setting its machinery in motion, of making the parts work efficiently and smoothly while they are yet untried and new.” Norwegian Nitrogen Products Co. v. United States, 288 U. S. 294, 315 (1933). And finally, and perhaps demanding particular weight, this constructtion has time and again been brought to the attention of the Joint Committee of Congress on Atomic Energy, which under § 202 of the Act, 42 TJ. S. C. § 2252, has a special duty during each session of Congress to “conduct hearings in either open or executive session for the purpose of receiving information concerning the development, growth, and state of the atomic energy industry,” and to oversee the operations of the AEC. See, e. g., Hearings on Development, Growth, and State of the Atomic Energy Industry, 84th Cong., 2d Sess., p. 106 (1956); Hearings on Development, etc., 85th Cong., 2d Sess., pp. 119-121 (1958) ; Hearings on Development, etc., 86th Cong., 2d Sess., pp. 103-109, 677-678 (1960); Hearings on Development, etc., 87th Cong., 1st Sess., pp. 29-32 (1961); Hearings on *409Governmental Indemnity for Private Licensees and AEC Contractors Against Reactor Hazards, 84th Cong., 2d Sess., pp. 62-65 (1956); A Study of AEC Procedures and Organization in the Licensing of Reactor Facilities, 85th Cong., 1st Sess., pp. 11-14, 100-108 (Joint Comm. Print 1957). No change in this procedure has ever been suggested by the Committee, although it has on occasion been critical of other aspects of the PRDC proceedings not before us. It may often be shaky business to attribute significance to the inaction of Congress, but under these circumstances, and considering especially the peculiar responsibility and place of the Joint Committee on Atomic Energy in the statutory scheme, we think it fair to read this history as a de jacto acquiescence in and ratification of the Commission’s licensing procedure by Congress. Cf., e. g., Ivanhoe Irrig. Dist. v. McCracken, 357 U. S. 275, 292-294 (1958); Brooks v. Dewar, 313 U. S. 354, 360-361 (1941). This same procedure has been used in each of the nine instances in which The Commission has granted a provisional construction permit for a developmental nuclear power reactor, e. g., Yankee Atomic Elec. Co., CPPR-5 (AEC 1957), and we hold that it was properly used in this case.

It is plain that the statute and regulations, as so construed and applied, were complied with fully. The Commission did not, as respondents’ argument seems at times to suggest, find merely that the construction of the reactor would present no safety problem. The Commission’s opinion and findings clearly were deeply concerned about the prospective safety of operation of the proposed reactor. Admitting that on the basis of the facts before it it was unable to make a definitive finding of safety, the Commission nevertheless found — and respondents do not deny that the finding was supported by substantial evidence — that it had information sufficient to provide *410reasonable assurance that the general type of reactor proposed could be operated without undue risk to the health and safety of the public. Its Finding 22, which we have quoted, was in the very words of Reg. 50.35, except for the insertion of the phrase, “for the purposes of this provisional construction permit.” This phrase was merely declaratory of the nature of the proceeding before the Commission, and in no way denigrated the finding as to safety of operation.

Respondents contend nevertheless that their construction of the statute is compelled by the legislative history. Since the Court of Appeals relied heavily on this history, we have studied it carefully. Two incidents are cited in particular. First, the Joint Committee stated in its report on the bill which became the Atomic Energy Act of 1954, and which when reported contained §§ 182 and 185 in substantially their present shape, that “[s]ection 185 . . . requires the issuance of a license if the construction is carried out in accordance with the terms of the construction permit.” S. Rep. No. 1699, 83d Cong., 2d Sess., p. 28 (1954); H. R. Rep. No. 2181, 83d Cong., 2d Sess., p. 28 (1954). The best we can say about this statement, with all deference, is that it must have been inadvertent. Witnesses who appeared before the Joint Committee at the hearings on the bill had made the very complaint that under the words of the bill as proposed a company might invest large sums in construction of a reactor, and then be denied the right to operate it. This situation, they claimed, was unfair, and would substan-. tially discourage the private investment in the field of atomic power which it was one of thp bill’s major purposes to stimulate. See Hearings before the Joint Committee on Atomic Energy on the Bill to Amend the Atomic Energy Act of 1946, 83d Cong., 2d Sess., Pt. I., pp. 113, 119 (statement of Paul W. McQuillen, representing *411the Dow Chemical-Detroit Edison and Associates atomic power development project, predecessors of PRDC); pp. 226-227 (statement of E. H. Dixon, chairman of the Committee on Atomic Power of the Edison Electric Institute and president of Middle-South Utilities, Inc.); p. 417 (statement of the Special Committee on Atomic Energy of the Association of the Bar of the City of New York). In spite of these pleas, however, the bill was unchanged. Industry spokesmen renewed the argument the next year when they sought unsuccessfully to have § 185 amended. Hearings on Development, etc., 84th Cong., 1st Sess., pp. 258, 261 (1955). Even a glance at § 185 suffices to show that issuance of a construction permit does not make automatic the later issuance of a license to operate. For that section sets forth three conditions, in addition to the completion of the construction, which must be met before an operating license is granted: (1) filing of any additional information necessary to bring the application up to date — information which will necessarily in this case include detailed safety data concerning the final design of petitioner’s reactor; (2) a finding that the reactor will operate in accordance with the act and regulations — i. e., that the safety and health of the public will be adequately protected — and with the construction permit itself, which is expressly conditioned upon a full investigation and finding of safety before operation is permitted; and (3) the absence of any good cause why the granting of a license to operate would not be in accordance with the Act — e. g., a showing by respondent unions, who will have full rights to appear and contest the issuance of an operating license, that the reactor may not be reasonably safe.

Respondents rely more heavily on another event during the debates on this bill on the floor of the Senate. Senator Humphrey, an opponent of the bill, expressed a *412desire that it be made clear that “the construction permit is equivalent to a license,” and that “the revised section 182 on license application . . . appl[ies] directly to construction permits.” 100 Cong. Rec. 12014 (July 26,1954). Senator Hickenlooper, floor manager of the bill and the ranking Senate member of the Joint Committee on Atomic Energy, indicated that he agreed with this construction of §§ 182 and 185. Senator Humphrey wanted these matters made clear because he feared that otherwise a construction permit could be easily obtained and substantial investment made in construction, and then the Commission would feel obliged, perhaps under pressure, to issue an operating license in order that this investment should not go to waste. The language used in the exchange between Senators Humphrey and Hickenlooper is susceptible, if read broadly and out of context, of the construction which respondents attribute to it, namely, that no § 185 construction permit may be issued unless the Commission has made the same safety-of-operation finding which it must make under § 182a before allowing actual operation. But the context of the exchange makes it clear that no such implication was intended by the participants. Senator'Humphrey’s statements were made during the consideration of an amendment which he had himself proposed on July 16. This amendment would have ádded the following clause to the end of § 185:

“and no construction permit shall be issued by the Commission until after the completion of the procedures established by section 182 for the consideration of applications for licenses under this act.”

Upon being assured by Senator Hickenlooper that an earlier amendment which Senator Hickenlooper himself had offered to § 189 took care of the problem, Senator Humphrey withdrew his proposal. This amendment to *413§ 189, which was adopted, was concerned solely with hearings and judicial review. Plainly Senator Humphrey’s concern was not with the substantive safety findings necessary to the issuance of a construction permit, but rather with the procedural safeguards with which that issuance should, in his opinion, be surrounded. The reference to the application of § 182 to construction permits was made not with § 182a in mind — that subsection sets out the substantive safety standard for the issuance of an operating license — but rather with a view to the application of § 182b, about which Senator Humphrey particularly asked Senator Hickenlooper during the exchange on the floor referred to, and which merely provides that notice of a license application must be published and given to any appropriate regulatory agencies, a procedural requirement which was fully satisfied in this case. This interpretation of the meaning of Senator Humphrey’s remarks is borne out by a statement of Representative Holifield, who, together with Representative Price, had dissented from the favorable report of the Joint Committee, precisely because, inter alia, under the bill as reported a construction permit did not have to be preceded by the same procedures as an operating license. See S. Rep. No. 1699, 83d Cong., 2d Sess., p. 123 (1954); H. R. Rep. No. 2181, 83d Cong., 2d Sess., p. 123 (1954). Representative Price wanted the same amendment added to § 185 which Senator Humphrey proposed, and he characterized this amendment as necessary to ensure “that the same procedural safeguards in the case of licenses be applied to construction permits.” 100 Cong. Rec. 10959 (July 19,1954). We think, therefore, that Senator Humphrey’s statement referred only to procedural prerequisites of construction permits, and had nothing to do with the substantive safety considerations which this case involves. If there were any doubt about this matter, the *414consistent administrative practice, made known to Congress many times and never disturbed by it, would dictate this conclusion.

The Court of Appeals put forward as an alternative basis for its decision the holding that under the law the Commission may not authorize the construction of a reactor near a large population center without “compelling reasons” for doing so, 108 U. S. App. D. C., at 103-104, 280 F. 2d, at 651-652, and that no such reasons had been found by the AEC in this case. It is not clear whether respondents have abandoned that contention in this Court, and it is likewise uncertain whether they ever presented it to the Commission, a step which would ordinarily be a prerequisite to its consideration by the Court of Appeals. In any event, the position is without merit. The statute and regulations say nothing about “compelling reasons.” Of course Congress (and the Commission, too, for that matter) had the problem of safety uppermost in mind, and of course that problem is most acute when a reactor, potentially dangerous, is located near a large city. But the Commission found reasonable assurance, for present purposes, that the reactor could be safely operated at the proposed location, and that is enough to. satisfy the requirements of law. The Commission recognized that the site and all its properties are among the most important ingredients of a finding of safety vel non. It considered the site along with all the other relevant data. There is no warrant in the statute for setting aside the Commission's conclusion.

We hold, therefore, that the Court of Appeals erred in setting aside the order of the AEC continuing PRDC's provisional construction permit in effect. We deem it appropriate to add a few words concerning the fears of nuclear disaster which respondents so urgently place before us. The respondents’ argument is tantamount to *415an insistence that the Commission cannot be counted on, when the time comes to make a definitive safety finding, wholly to exclude the consideration that PRDC will have made an enormous investment. The petitioners concede that the Commission is absolutely denied any authority to consider this investment when acting upon an application for a license for operation. PRDC has been on notice long since that it proceeds with construction at its own risk, and that all its funds may go for naught. With its eyes open, PRDC has willingly accepted that risk, however great. No license to operate may be issued to PRDC until a full hazards report has been filed, until the AEC’s Advisory Committee on Reactor Safeguards makes a full investigation and public report on safety to the Commission, until the Commission itself, after notice and hearings at which respondents, if they desire, may be heard, has made the safety-of-operation finding required by § 182a and Reg. 50.35, and until the other requirements of § 185 have been met. It may be that an operating license will never be issued. If one is, that will not be the end of the matter. The respondents may have judicial review. Moreover, the Commission’s responsibility for supervision of PRDC continues. For, under Reg. 50.57, 10 CFR § 50.57, operation at full power (100,000 electric kilowatts) will not be permitted until several steps of gradually increasing operation have been successfully mastered, with a full public hearing at each step, and no further advance permitted without the AEC’s being fully satisfied that a step-up will meet the high safety standards imposed by law. This is the multi-step scheme which Congress and the Commission have devised to protect the public health and safety. We hold that the actions of the Commission up to now have been within the Congressional authorization. We cannot assume that the Commission will exceed its powers, or that these *416many safeguards to protect the public interest will not be fully effective.

Accordingly, the judgment is reversed and the causes are remanded to the Court of Appeals for further proceedings consistent with this opinion.

Reversed and remanded.

Mr. Justice Douglas,

with whom Mr. Justice Black concurs, dissenting.

The only requirement in the Act for a finding that the facilities involved here “will provide adequate protection to the health and safety of the public” is found in § 182 which is headed “License Applications.” 1 By the terms of § 185 a construction permit is, apart from the requirements of § 185, “deemed to be a ‘license.’ ” 2 Section 185 governs applications for construction permits. It has no separate or independent standards for safety, no specific requirement for a finding on “safety.” If the facility is finished and will operate “in conformity with” the Act, the license issues “in the absence of any good cause being shown to the Commission why the granting of a license would not be in accordance with the provisions of” the Act. As the Committee Report stated, “Section 185 . . . requires the issuance of a license if the construction is carried out in accordance with the terms of the construction permit.” 3 In other words, the finding on “safety,” if it is to be made (as it assuredly must be), must be made at the time the construction permit is issued or not at all.

While in the present case the Commission “finds reasonable assurance in the record, for the purposes of this provisional construction permit,” that the facility can be operated “without undue risk to the health and safety of *417the public,” it also finds that “It has not been positively-established” that a facility of this character “can be operated without a credible possibility of releasing significant quantities of fission products to the environment.” The Commission added that there was “reasonable assurance” before the date when the facility went into operation that research and investigation would definitely establish “whether or not the reactor proposed by Applicant can be so operated.”

Plainly these are not findings that the “safety” standards have been met. They presuppose — contrary to the premise of the Act — that “safety” findings can be made after construction is finished. But when that point is reached, when millions have been invested, the momentum is on the side of the applicant, not on the side of the public. The momentum is not only generated by the desire to salvage an investment. No agency wants to be the architect of a “white elephant.” Congress could design an Act that would give a completed structure that momentum. But it is clear to me it did not do so.

When this measure was before the Senate, Senator Humphrey proposed an amendment that read, “no construction permits shall be issued by the Commission until after the completion of the procedures established by section 182 for the consideration of applications for licenses under this act.” 4 That amendment would plainly have made the present findings inadequate, for they leave the issue of “safety” wholly in conjecture and unresolved.

Senator Humphrey explained his amendment as follows: 5

“The purpose of the amendment when it was prepared was to make sure that the construction of a facility was not permitted prior to the authorization *418of a license, because had that been done what it would have amounted to would be getting an investment of a substantial amount of capital, which surely would have been prejudicial in terms of the Commission issuing the license. In other words, if the Commission had granted the construction permit for some form of nuclear reactor, and then the question of a license was not fully resolved, surely there would have been considerable pressure, and justifiably so, for the Commission to have authorized the license once it had authorized the permit for construction.

“The chairman of the committee tells me he has modified certain sections by the committee amendments to the bill, of which at that time I was not aware. The chairman indicates to me that under the terms of the bill, as amended, the construction permit is equivalent to a license. In other words, as I understand, under the bill a construction permit cannot be interpreted in any other way than being equal to or a part of the licensing procedure. Is that correct?”

His question was answered by Senator Hickenlooper, who was in charge of the bill: 6

“A license and a construction permit are equivalent. They are the same thing, and one cannot operate until the other is granted.

“The same is true with reference to hearings. Therefore, we believe, and we assure the Senator, that the amendment is not essential to the problem which he is attempting to reach.”

Senator Humphrey then asked if § 182 applied “directly to construction permits.” 7 Senator Hickenlooper *419replied “Yes.” 8 Senator Humphrey accordingly withdrew his amendment.9

This legislative history makes clear that the time when the issue of “safety” must be resolved is before the Commission issues a construction permit. The construction given the Act by the Commission (and today approved) is, with all deference, a light-hearted approach to the most awesome, the most deadly, the most dangerous process that man has ever conceived.10

APPENDIX TO OPINION OF MR. JUSTICE DOUGLAS.

Section 182a provides in relevant part:

“License Applications.^—

“a. Each application for a license hereunder shall be in writing and shall specifically state such information as the Commission, by rule or regulation, may determine to be necessary to decide such of the technical and financial qualifications of the applicant, the character of the applicant, the citizenship of the applicant, or any other qualifications of the applicant as the Commission may deem appropriate for the license. In connection with applications for licenses to operate production or utilization facilities, the applicant shall state such technical specifications, including information of the amount, kind, *420and source of special nuclear material required, the place of the use, the specific characteristics of the facility, and such other information as the Commission may, by rule or regulation, deem necessary in order to enable it to find that the utilization or production of special nuclear material will be in accord with the common defense and security and will provide adequate protection to the health and safety of the public. Such technical specifications shall be a part of any license issued.”

Section 185 provides:

“Construction Permits. — All applicants for licenses to construct or modify production or utilization facilities shall, if the application is otherwise acceptable to the Commission, be initially granted a construction permit. The construction permit shall state the earliest and latest dates for the completion of the construction or modification. Unless the construction or modification of the facility is completed by the completion date, the construction permit shall expire, and all rights thereunder be forfeited, unless upon good cause shown, the Commission extends the completion date. Upon the completion of the construction or modification of the facility, upon the filing of any additional information needed to bring the original application up to date, and upon finding that the facility authorized has been constructed and will operate in conformity with the application as amended and in conformity with the provisions of this Act and of the rules and regulations of the Commission, and in the absence of any good cause being shown to the Commission why the granting of a license would not be in accordance with the provisions of this Act, the Commission shall thereupon issue a license to the applicant. For all other purposes of this Act, a construction permit is deemed to be a 'license.’ ”

THOMAS, Circuit Judge.

This case presents the question, inter alia, as to whether the likely environmental consequences of a potential terrorist attack on a nuclear facility must be considered in an environmental review required under the National Environmental Policy Act. The United States Nuclear Regulatory Commission (“NRC”) contends that the possibility of a terrorist attack on a nuclear facility is so remote and speculative that the potential consequences of such an attack need not be considered at all in such a review. The San Luis Obispo Mothers for Peace and other groups disagree and petition for review of the NRC’s approval of a proposed Interim Spent Fuel Storage Installation. We grant the petition in part and deny it in part.

I

The NRC is an independent federal agency established by the Energy Reorganization Act of 1974 to regulate the civilian use of nuclear materials. Intervenor Pacific Gas and Electric Company (“PG & E”) filed an application with the NRC under 10 C.F.R. Part 72 for a license to construct and operate an Interim Spent Fuel Storage Installation (“Storage Instal*1020lation” or “ISFSI”) at PG & E’s Diablo Canyon Power Plant (“Diablo Canyon”) in San Luis Obispo, California. The NRC granted the license. The question presented by this petition for review is whether, in doing so, the NRC complied with federal statutes including the National Environmental Policy Act of 1969 (“NEPA”), 42 U.S.C. §§ 4821—4437, the Atomic Energy Act of 1954 (“AEA”), 42 U.S.C. §§ 2011-2297g, and the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 551-706.

NEPA establishes a “national policy [to] encourage productive and enjoyable harmony between man and his environment,” and was intended to reduce or eliminate environmental damage and to promote “the understanding of the ecological systems and natural resources important to” the United States. Dept. of Transp. v. Pub. Citizen, 541 U.S. 752, 756, 124 S.Ct. 2204, 159 L.Ed.2d 60 (2004) (quoting 42 U.S.C. § 4321). The Supreme Court has identified NEPA’s “twin aims” as “placing] upon an agency the obligation to consider every significant aspect of the environmental impact of a proposed action!, and] ensuring] that the agency will inform the public that it has indeed considered environmental concerns in its deci-sionmaking process.” Baltimore Gas & Elec. Co. v. Natural Res. Def. Council, Inc., 462 U.S. 87, 97, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983).

Rather than mandating particular results, NEPA imposes on federal agencies procedural requirements that force consideration of the environmental consequences of agency actions. Pub. Citizen, 541 U.S. at 756, 124 S.Ct. 2204. At NEPA’s core is the requirement that federal agencies prepare an environmental impact statement (“EIS”), or:

include in every recommendation or report on proposals for legislation and other major Federal actions significantly affecting the quality of the human environment, a detailed statement by the responsible official on — (i) the environmental impact of the proposed action, (ii) any adverse environmental effects which cannot be avoided should the proposal be implemented, (iii) alternatives to the proposed action, (iv) the relationship between local short-term uses of man’s environment and the maintenance and enhancement of long-term productivity, and (v) any irreversible and irretrievable commitments of resources which would be involved in the proposed action should it be implemented.

Id. at 757, 124 S.Ct. 2204 (quoting 42 U.S.C. § 4332(2)(C)).

As an alternative to the EIS, an agency may prepare a more limited environmental assessment (“EA”) concluding in a “Finding of No Significant Impact” (“FONSI”), briefly presenting the reasons why the action will not have a significant impact on the human environment. Id. at 757-58, 124 S.Ct. 2204 (citing 40 C.F.R. §§ 1501.4(e), 1508.13). If, however, the EA does not lead to the conclusion that a FONSI is warranted, the agency remains obligated to prepare an EIS. Id. at 757, 124 S.Ct. 2204.

While NEPA requires the NRC to consider environmental effects of its decisions, the AEA is primarily concerned with setting minimum safety standards for the licensing and operation of nuclear facilities. The NRC does not contest that the two statutes impose independent obligations, so that compliance with the AEA does not excuse the agency from its NEPA obligations. The AEA lays out the process for consideration of the public health and safety aspects of nuclear power plant licensing, and requires the NRC to determine whether the licensing and operation of a proposed facility is “in accord with the *1021common defense and security and will provide adequate protection to the health and safety of the public.” 42 U.S.C. § 2232(a).

The NRC is not, however, required to make this determination without assistance; federal law provides a framework for hearings on material issues that interested persons raise by specific and timely petition. 42 U.S.C. § 2239(a); 10 C.F.R. §§ 2.308-.348; 5 U.S.C. §§ 551-706. The initial hearing is held before a three-person Atomic Safety and Licensing Board (“Licensing Board”). 10 C.F.R. § 2.321. The Licensing Board’s findings and decision constitute the agency’s initial determination, although a party may file a petition for review with the Commission within 15 days of the Licensing Board’s decision. 10 C.F.R. § 2.341. If the petition is granted, the Commission specifies the issues to be reviewed and the parties to the review proceedings, 10 C.F.R. § 2.341(c)(1), and renders a final decision. 10 C.F.R. § 2.344. A party may then petition this court for review of the Commission’s final decision. 28 U.S.C. § 2344.

II

With this general statutory background, we turn to the facts underlying the petition for review. On December 21, 2001, PG & E applied to the NRC pursuant to 10 C.F.R. Part 72 for a license to construct and operate a Storage Installation at Dia-blo Canyon. The Storage Installation would permit the necessary and on-site storage of spent fuel, the byproduct of the two nuclear reactors at that site. PG & E expects to fill its existing spent fuel storage capacity at Diablo Canyon sometime this year. Therefore, unless additional spent fuel storage capacity is created, the Diablo Canyon reactors cannot continue to function beyond 2006.

PG & E proposes to build a dry cask storage facility. The basic unit of the storage system is the Multi-Purpose Canister (“Canister”), a stainless steel cylinder that is filled with radioactive waste materials and welded shut. The Canisters are loaded into concrete storage overpacks that are designed to permit passive cooling via the circulation of air. The storage casks, or the filled Canisters loaded into overpacks, are then placed on one of seven concrete pads. The Storage Installation would house a total of 140 storage casks, 2 more than the 138 projected to be required for storage of spent fuel generated at Dia-blo Canyon through 2025.

On April 22, 2002, the NRC published a Notice of Opportunity for Hearing. Under the regulatory scheme, interested parties could then request a hearing or petition for leave to intervene. 10 C.F.R. § 2.309(a). A written hearing request, which must contain the contentions the party wants litigated at the hearing, will be granted if the petitioner has standing, and has posed at least one admissible contention.1 Id.

On July 19, 2002, the San Luis Obispo Mothers for Peace, a non-profit corporation concerned with Diablo Canyon’s local impact, the Sierra Club, a non-profit corporation concerned with national environmental policy, and Peg Pinard, an indi*1022vidual citizen, (collectively “Petitioners”) submitted a hearing request and a petition to intervene, asserting contentions for admission.

In Licensing Board Proceeding LBP-02-23, 56 NRC 413 (“LBP 02-23”), the Atomic Safety and Licensing Board addressed the admissibility of the July 19 petition’s five Technical and three Environmental Contentions.2 One Technical Contention, TC-1, dealing with the state of PG & E’s finances, was deemed admissible; the acceptance of at least one contention meant that the petition was granted. Although the Licensing Board deemed two Environmental Contentions, EC-1, dealing with the failure to address environmental impacts of terrorist or other acts of malice or insanity, and EC-3, dealing with the failure to evaluate environmental impacts of transportation of radioactive materials3 inadmissible, the Licensing Board nonetheless referred the final ruling as to the admissibility of these two contentions to the NRC, “in light of the Commission’s ongoing ‘top to bottom’ review of the agency’s safeguards and physical security programs.” 56 NRC at 448.

ín a memorandum and order, CLI-03-1, 57 NRC 1 (“CLI 03-01”), the NRC accepted the Licensing Board’s referral of its decision to reject the environmental contentions related to terrorism. Although the Commission affirmed the Licensing Board’s rejection of the contentions, it based its decision on a different rationale. The NRC relied on four prior decisions in which it held that the NEPA does not require a terrorism review.4 These decisions, most particularly Private Fuel Storage, CLI-02-25, 56 NRC 340 (2002), outlined four reasons for this holding: (1) the possibility of terrorist attack is too far removed from the natural or expected consequences of agency action to require study under NEPA; (2) because the risk of a terrorist attack cannot be determined, the analysis is likely to be meaningless; (3) NEPA does not require a “worst-case” analysis; and (4) NEPA’s public process is not an appropriate forum for sensitive security issues. The NRC concluded:

Our decision today rests entirely on our understanding of NEPA and of what means are best suited to dealing with terrorism. Nonetheless, our conclusion *1023comports with the practical realities of spent fuel storage and the congressional policy to encourage utilities to provide for spent fuel storage at reactor sites pending construction of a permanent repository. Storage of spent fuel at commercial reactor sites offers no unusual technological challenges. Indeed, it has been occurring at Diablo Canyon for many years and will continue whether or not we license the proposed Installation.

57 NRC at 7.

In September of 2002, prior to the NRC’s decision on the first petition, Petitioners submitted a second petition, this time requesting suspension of the Storage Installation licensing proceeding pending comprehensive review of the adequacy of Diablo Canyon’s design and operation measures for protection against terrorist attack and other acts of malice or insanity. Unlike the July 19 petition, this one addressed security measures for the entire Diablo Canyon complex, not merely the Storage Installation. Petitioners explained that 10 C.F.R. § 2.335, which prohibits challenges to any NRC rule or regulation in an adjudicatory proceeding involving initial or renewal licensing, prevented the raising of contentions contesting the adequacy of NRC safety requirements protecting against terrorist or other malicious attacks on the entire complex in the July 19 Petition. Petitioners also stated that 10 C.F.R. § 72.32 prevented them from raising emergency planning contentions in the earlier petition. Thus, Petitioners insisted that the second petition “d[id] not constitute a request for rulemaking, nor ... for enforcement action,” and instead defined it, without reference to any particular hearing-granting provision of the regulations, as “a request for actions that are necessary to ensure that any licensing decision made by the Commission with respect to the proposed Diablo Canyon Installation complies with the Commission’s statutory obligations under the Atomic Energy Act.”

In a memorandum and order, CLI-02-23, 56 NRC 230 (“CLI 02-23”), the NRC denied the September 2002 petition. Because the petition did not, according to the NRC, “fit comfortably in any specific category, [the Commission] treat[ed] it as a general motion brought under the procedural requirements of 10 C.F.R. § 2.730.”5 In rejecting the petition, the Commission reasoned that by not suspending operating licenses at installations and power plants following the September 11, 2001 terrorist attacks, it had demonstrated its implicit conclusion that the continued operation of these facilities neither posed an imminent risk to the public health, nor was inimical to the common defense. Further, the Commission concluded that because it had already initiated a thorough review of its safeguards and physical security program, there was no reason to suspend the Diablo Canyon licensing proceeding to address the terrorism-related concerns raised by the Petitioners. It stated that “[t]here certainly is no reason to believe that any danger to public health and safety would result from mere continuation of this adjudicatory proceeding,” given that the proceeding was in its initial stages, that construction was not scheduled to begin for several years, and that the Petitioners would be able to comment on any changes in the rules resulting from the Commission’s ongoing review of terrorism-related matters if and when they were to occur.

In a memorandum and order, CLI-03-12, 58 NRC 185 (2003) (“CLI 03-02”), the NRC denied the petitions for agency review of the Licensing Board’s decisions *1024that “cumulatively, rejected challenges to [the PG & E] Installation application.” This denial thus became a final order, reviewable by this court on petition for review. 28 U.S.C. § 2344.

In October of 2003, the Spent Fuel Project Office of the NRC’s Office of Material Safety and Safeguards released its Environmental Assessment Related to the Construction and Operation of the Diablo Canyon Independent Spent Fuel Storage Installation. The 26-page document contains the NRC’s conclusion “that the construction, operation, and decommissioning of the Diablo Canyon Installation will not result in significant impact to the environment,” and therefore that “an [EIS] is not warranted for the proposed action, and pursuant to 10 C.F.R. [§ ] 51.31, a Finding of No Significant Impact is appropriate.”

The EA is not devoid of discussion of terrorist attacks. Indeed, the document contains the Commission’s response to a comment submitted by the California Energy Commission in response to an earlier draft that “there is no discussion in the EA of the potential destruction of the casks or blockage of air inlet ducts as the result of sabotage or a terrorist attack ... [nor is there] a description of how decisions are being made regarding the configuration, design and spacing of the casks, the use of berms, and the location of the ISFSI to minimize the vulnerability of the ISFSI to potential attack.” The NRC responded:

In several recent cases, ... the Commission has determined that an NRC environmental review is not the appropriate forum for the consideration of terrorist acts. The NRC staff considers the security of spent fuel as part of its safety review of each application for an ISFSI license. In addition to reviewing an ISFSI application against the requirements of 10 CFR Part 72, the NRC staff evaluates the proposed security plans and facility design features to determine whether the requirements in 10 CFR Part 73, “Physical Protection of Plants and Materials,” are met. The details of specific security measures for each facility are Safeguards Information, and as such, can not be released to the public.

The NRC has also initiated several actions to further ensure the safety of spent fuel in storage.- Additional security measures have been put in place at nuclear facilities, including ISFSIs currently storing spent fuel. These measures include increased security patrols, augmented security forces and weapons, additional security posts, heightened coordination with law enforcement and military authorities, and additional limitations on vehicular access. Also, as part of its comprehensive review of its security program, the NRC is conducting several technical studies to assess potential vulnerabilities of spent fuel storage facilities to a spectrum of terrorist acts. The results of these studies will be used to determine if revisions to the current NRC security requirements are warranted.

Petitioners argue that, in denying their petitions, the NRC violated the AEA, the APA, and NEPA. Although we reject the AEA and APA claims, we agree with Petitioners that the agency has failed to comply with NEPA. We have jurisdiction over those final orders of the NRC made reviewable by 42 U.S.C. § 2239, which includes final orders entered in licensing proceedings, under 28 U.S.C. § 2342(4).

Ill

We turn first to Petitioners’ AEA argument. Specifically, Petitioners argue that the NRC violated its regulations implementing the AEA, as well as the AEA’s *1025hearing provisions, when it denied Petitioners a hearing on whether NEPA required consideration of the environmental impact of a terrorist attack on the Storage Installation; they also argue that the NRC violated the AEA's hearing provisions in denying Petitioners a hearing on post-September 11th security measures for the entire Diablo Canyon complex. Both of these challenges fail.

A

The NRC did not violate the AEA or its implementing regulations when it failed to explain its rejection of Petitioners’ contentions by addressing each of their arguments. Nothing in the regulations or the AEA requires the NRC to provide such an explanation.

Section 189(a) of the AEA grants public hearing rights “upon the request of any person whose interest may be affected” by an NRC licensing proceeding. 42 U.S.C. § 2239. The NRC public hearing regulations, at 10 C.F.R. § 2.309, “promulgated pursuant to the AEC’s6 power to make, promulgate, issue, rescind, and amend such rules and regulations as may be necessary to carry out the purposes of’ the AEA, 42 U.S.C. § 2201(p), specify the procedures required of both petitioners and the NRC in making and deciding hearing petitions.

Petitioners correctly observe that the NRC, in its decision, did not discuss whether Petitioners satisfied the regulatory standard. They are mistaken, however, in their unsupported contention that this omission amounts to the agency’s failure to follow its own regulations and thus is “reversible error.” The regulations simply do not require the NRC to explain its decisions in any particular manner. Although the NRC regulations are specific and demanding in what they require of petitioners, they demand far less of the NRC in responding to a petition: the regulations require only a timely “decision.” See 10 C.F.R. § 2.714(f) (“Decision on request/petition. The presiding officer shall, within 45 days after the filing of answers and replies ... issue a decision on each request for hearing/petition to intervene.”). Because Petitioners do not claim that the NRC violated this requirement, we must reject this challenge.

B

The NRC’s denial of a hearing on whether NEPA requires consideration of the environmental effects of a terrorist attack on the Storage Installation did not violate the AEA’s hearing provisions.

Petitioners contend that the NRC relied on an improper ground in denying their request for a hearing on whether NEPA requires the Commission to consider the environmental impacts of terrorism— namely, the ground that it had determined in earlier decisions that NEPA imposes no such obligation. Thus, Petitioners do not challenge the substantive validity or coherence of those earlier opinions in making their AEA claim, but rather the reliance upon a prior determination of the merits in order to reject a petition presenting the same issues. As such, Sierra Club v. NRC, 862 F.2d 222 (9th Cir.1988), on which Petitioners rely, does not apply. In that case, the NRC rejected the petitioners’ contentions as lacking in reasonable specificity, and yet went on to analyze the merits of those supposedly unacceptable contentions. Id. at 228. Here, however, where the agency is rejecting the conten*1026tions as contrary to a prior decision, the “merits” and the reason for the inadmissibility of the contention collapse. Put differently, the NRC did not reach the merits of the petition as much as it assessed the issues raised against issues resolved by prior decisions. We hold that in doing so, the Commission complied fully with the AEA. To hold otherwise would unduly restrict the agency’s evaluation of hearing petitions, by requiring it to grant a hearing on issues it has already resolved whenever a petitioner claims to have new evidence. We can find, and Petitioners point to, nothing in the AEA that would require this result.

C

The NRC’s denial of a hearing on security measures for Diablo Canyon as a whole also did not violate the AEA. Petitioners argue that the AEA requires the NRC to grant petitioners a hearing on all issues of material fact, including the security of the entire Diablo Canyon complex. Petitioners therefore conclude, citing Union of Concerned, Scientists v. NRC, 735 F.2d 1437 (D.C.Cir.1984), that the NRC violated the AEA when it denied a hearing on that issue.

Petitioners’ argument misreads Union of Concerned Scientists, in which the D.C. Circuit held only that the agency cannot by rule presumptively eliminate a material issue from consideration in a hearing petition. Union of Concerned Scientists requires the agency to consider a petition; it does not require that the agency grant it.

The NRC in CLI 02-23 did not deny that security requirements for the entire complex might need to be upgraded, but rather maintained that a licensing proceeding hearing (and one regarding an installation, not the entire complex) was not the correct forum in which to address the issue. The Commission directed Petitioners to participate in a rulemaking or to raise their concerns in a hearing then pending before the Licensing Board. Petitioners contend that these alternative fora are illusory, and that rejection of their petition amounted to the denial of any opportunity to participate in the consideration of post-9/11 security measures for the Diablo Canyon complex.

Petitioners argue “[i]f the NRC were going to resolve Petitioners’ concerns that grossly inadequate security made the Dia-blo Canyon facility vulnerable to terrorist attacks generieally, through a rulemaking, such a rulemaking would have been initiated as a result of the ‘comprehensive security review’ undertaken by the NRC.” Thus, Petitioners argue that it would have been futile to submit a rulemaking petition. This argument must fail, as Petitioners did not use the available procedures for initiating a rulemaking. Petitioners cannot complain that NRC failed to institute a rulemaking they never requested.

Given that rulemaking may have been an avenue for Petitioners’ participation, had they chosen to pursue it, their argument that they had no forum in which to raise them contentions loses its force. However, even were Petitioners correct in their assertion that they were unfairly denied the opportunity to participate in a rulemaking proceeding, the argument that the Licensing Board hearing was similarly illusory would fail. In fact, Petitioners were attempting to use the present Storage Installation licensing proceeding as a means of launching a much broader challenge to the Diablo Canyon complex. The NRC correctly observes that a petition alleging that existing NRC regulations are “grossly inadequate to protect against terrorist attack, and therefore must be supplemented by additional requirements” cannot in fact be raised before the Licens*1027ing Board, which cannot hear challenges to NRC rules. The limited scope of licensing proceedings does not, however, amount to the arbitrary denial of a forum, as Petitioners claim. While Petitioners could have raised site-specific issues “relating to the ‘common defense and security”’ that were not controlled by existing rules or regulations to the Licensing Board, they are not entitled to expand those proceedings to include the entire complex, and issues already covered by agency rules.

D

In short, the NRC did not violate the AEA in denying the petitions for a hearing. Neither the AEA nor its implementing regulations required the NRC to grant Petitioners a hearing on whether NEPA required a consideration of the environmental impact of a terrorist attack on the Storage Installation or the security measures adopted for the entire Diablo Canyon complex.

IV

The NRC’s reliance on its own pri- or opinions in its decision in this case does not violate the APA’s notice and comment provisions. Petitioners argue that the decisions in CLI 03-01 and PFS amount to the announcement “of a general policy of refusing to consider the environmental impacts of terrorist attacks in Environmental Impact Statements.” Petitioners rely on Mada-Luna v. Fitzpatrick, 813 F.2d 1006, 1014 (9th Cir.1987) to claim that this policy depends on factual determinations not found subsequent to an evidentiary proceeding, and constitutes a “binding substantive norm,” the promulgation of which, without a public hearing, violates the APA notice and comment provisions contained in 5 U.S.C. §§ 553(b), (c).7 The flaw in Petitioners’ argument is the mistaken assertion that the NRC’s decisions were factual and not legal. If the NRC’s conclusion that terrorism need not be examined under NEPA were factual, then Petitioners would be correct that its determination would have to comply with APA rulemak-ing requirements, including notice and comment, or else the agency would have to permit petitioners to challenge it in every proceeding where it was disputed.

That NEPA does not require consideration of the environmental impacts of terrorism is a legal, and not a factual, conclusion. Cf. Greenpeace Action v. Franklin, 14 F.3d 1324, 1331 (9th Cir.1993) (reasoning that a challenge to the adequacy of an EA turned on factual, not legal, principles where both NEPA’s applicability and the requirements it imposed were uncontested); see also Alaska Wilderness Recreation & Tourism Ass’n v. Morrison, 67 F.3d 723, 727 (9th Cir.1995) (noting that although “challenges to agency actions which raise predominantly legal, rather than technical questions, are rare,” the court was there required to address “just such a challenge”). Petitioners’ analysis is therefore inapposite. The agency has the discretion to use adjudication to establish a binding legal norm. See Sec. & Exch. Comm’n v. Chenery, 332 U.S. 194, 199-203, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947) (“[T]he choice made between proceeding by general rule or by individual, ad hoc litigation, is one that lies primarily in the informed discretion of the administrative agency.”). We therefore agree with the NRC’s char*1028acterization in its brief to this court: having come to the legal conclusion that NEPA does not require consideration of the environmental consequences of terrorist attacks, “[w]hen petitioners in this case presented a proposed contention seeking an EIS that analyzed the impacts of possible terrorist acts at the proposed Diablo Canyon Installation, the NRC reasonably concluded that this request was sufficiently similar to the request in PFS to justify the application of that decision here.”

V

Although we hold that the agency did not violate the APA when it relied on the prior resolution of a legal issue through adjudication, we come to a different conclusion as to that determination’s compliance with NEPA. Because the issue whether NEPA requires consideration of the environmental impacts of a terrorist attack is primarily a legal one, we review the NRC’s determination that it does not for reasonableness. See Alaska Wilderness Recreation & Tourism Ass’n, 67 F.3d at 727 (reviewing predominately legal issue for reasonableness because “it makes sense to distinguish the strong level of deference we accord an agency in deciding factual or technical matters from that to be accorded in disputes involving predominately legal questions”); Ka Makani‘O Kohala Ohana, Inc. v. Water Supply, 295 F.3d 955, 959 n. 3 (9th Cir.2002) (“Because this case involved primarily legal issues ... based on undisputed historical facts, we conclude that the ‘reasonableness’ standard should apply to this case.”).

Here, the NRC decided categorically that NEPA does not require consideration of the environmental effects of potential terrorist attacks. In making this determination, the NRC relied on PFS, where it “considered] in some detail the legal question whether NEPA requires an inquiry into the threat of terrorism at nuclear facilities.” 56 NRC 340, 343 (2002). In that case, intervenor State of Utah filed a contention claiming that the September 11 terrorist attacks “had materially changed the circumstances under which the Board had rejected previously proffered terrorism contentions by showing that a terrorist attack is both more likely and potentially more dangerous than previously thought.” Id. at 345. The NRC concluded that even following the September 11th attacks, NEPA did not impose such a requirement, reasoning:

In our view, an EIS is not an appropriate format to address the challenges of terrorism. The purpose of an EIS is to inform the decisionmaking authority and the public of a broad range of environmental impacts that will result, with a fair degree of likelihood, from a proposed project, rather than to speculate about ‘worst-case’ scenarios and how to prevent them.

Id. at 347.

The NRC determined that four grounds “eut[] against using the NEPA framework” to consider the environmental effects of a terrorist attack: (1) the possibility of a terrorist attack is far too removed from the natural or expected consequences of agency action; (2) because the risk of a terrorist attack cannot be determined, the analysis is likely to be meaningless; (3) NEPA does not require a “worst-case” analysis; and (4) NEPA’s public process is not an appropriate forum for sensitive security issues. Id. at 348. We review each of these four grounds for reasonableness, and conclude that these grounds, either individually or collectively, do not support the NRC’s categorical refusal to consider the environmental effects of a terrorist attack.

*1029A

The Commission relied first on finding that the possibility of a terrorist attack is too far removed from the natural or expected consequences of agency action. Id. at 347. Section 102 of NEPA requires federal agencies to prepare “a detailed statement ... on the environmental impact” of any proposed major federal action “significantly affecting the quality of the human environment.” 42 U.S.C. § 4332(1)(C)(i). The question thus becomes whether a given action “significantly affects” the environment.

The NRC claims that the appropriate analysis of Section 102 is that employed by the Supreme Court in Metropolitan Edison Co. v. People Against Nuclear Energy, 460 U.S. 766, 773, 103 S.Ct. 1556, 75 L.Ed.2d 534 (1983). In Metropolitan Edison, the Court noted that “[t]o determine whether Section 102 requires consideration of a particular effect, we must look to the relationship between that effect and the change in the physical environment caused by the major federal action at issue,” looking for “a reasonably close causal relationship ... like the familiar doctrine of proximate cause from tort law.” 460 U.S. at 774, 103 S.Ct. 1556. The Commission claims that its conclusion that the environmental impacts of a possible terrorist attack on an NRC-licensed facility is beyond a “reasonably close causal relationship” was a reasonable application of this “proximate cause” analogy.

The problem with the agency’s argument, however, is that Metropolitan Edison and its proximate cause analogy are inapplicable here. In Metropolitan Edison, the petitioners argued that NEPA required the NRC to consider the potential risk of psychological damage upon reopening the Three Mile Island nuclear facilities to those in the vicinity. Noting that NEPA is an environmental statute, the Supreme Court held that the essential analysis must focus on the “closeness of the relationship between the change in the environment and the ‘effect’ at issue.” 460 U.S. at 772, 103 S.Ct. 1556.

The appropriate analysis is instead that developed by this court in No GWEN Alliance v. Aldridge, 855 F.2d 1380 (9th Cir.1988). In No GWEN, the plaintiffs argued that NEPA required the Air Force to consider the threat of nuclear war in the implementation of the Ground Wave Emergency Network (“GWEN”). We held “that the nexus between construction of GWEN and nuclear war is too attenuated to require discussion of the environmental impacts of nuclear war in an[EA] or [EIS].” 855 F.2d at 1386.

The events at issue here, as well as in Metropolitan Edison and No GWEN, form a chain of three events: (1) a major federal action; (2) a change in the physical environment; and (3) an effect. Metropolitan Edison was concerned with the relationship between events 2 and 3 (the change in the physical environment, or increased risk of accident resulting from the renewed operation of a nuclear reactor, and the effect, or the decline in the psychological health of the human population). The Court in Metropolitan Edison explicitly distinguished the case where the disputed relationship is between events 1 and 2: “we emphasize that in this case we are considering effects caused by the risk of accident. The situation where an agency is asked to consider effects that will occur if a risk is realized, for example, if an accident occurs ... is an entirely different case.” Id. at 775 n. 9, 103 S.Ct. 1556. In No GWEN, we followed the Court’s admonition and, in addressing the relationship between events 1 and 2, we held that the Metropolitan Edison analysis did not apply “because it discusse[d] a different type of causation than that at issue in this case *1030... -[which] require[d] us to examine the relationship between the agency action and a potential impact on the environment.” Id. at 1386. No GWEN relied on our decision in Warm Springs Dam Task Force v. Gribble, 621 F.2d 1017, 1026 (9th Cir.1980), which held that “an impact statement need not discuss remote and highly speculative consequences.” Applying that standard to the plaintiffs’ claims that the military GWEN system’s installation would “increase the probability of nuclear war,” and “that GWEN would be a primary target in a nuclear war,” we held both propositions to be “remote and highly speculative,” and, therefore, NEPA did not require their consideration.

In the present case, as in No GWEN, the disputed relationship is between events 1 and 2 (the federal act, or the licensing of the Storage Installation, and the change in the physical environment, or the terrorist attack). The appropriate inquiry is therefore whether such attacks are so “remote and highly speculative” that NEPA’s mandate does not include consideration of their potential environmental effects.

The NRC responds by simply declaring without support that, as a matter of law, “the possibility of a terrorist attack ... is speculative and simply too far removed from the natural or expected consequences of agency action to require a study under NEPA.” 56 NRC at 349. In doing so, the NRC failed to address Petitioners’ factual contentions that licensing the Storage Installation would lead to or increase the risk of a terrorist attack because (1) the presence of the Storage Installation would increase the probability of a terrorist attack on the Diablo Canyon nuclear facility, and (2) the Storage Installation itself would be a primary target for a terrorist attack. We conclude that it was unreasonable for the NRC to categorically dismiss the possibility of terrorist attack on the Storage Installation and on the entire Diablo Canyon facility as too “remote and highly speculative” to warrant consideration under NEPA.

In so concluding, we also i*ecognize that the NRC’s position that terrorist attacks are “remote and highly speculative,” as a matter of law, is inconsistent with the government’s efforts and expenditures to combat this type of terrorist attack against nuclear facilities. In the PFS opinion, the NRC emphasized the agency’s own post-September 11th efforts against the threat of terrorism:

At the outset, however, we stress our determination, in the wake of the horrific September 11th terrorist attacks, to strengthen security at facilities we regulate. We currently are engaged in a comprehensive review of our security regulations and programs, acting under our AEA-rooted duty to protect “public health and safety” and the “common defense and security.” We are reexamining, and in may cases have already improved, security and safeguards matters such as guard force size, physical security exercises, clearance requirements and background investigations for key employees, and fitness-for-duty requirements. More broadly, we are rethinking the NRC’s threat assessment framework and design basis threat. We also are reviewing our own infrastructure, resources, and communications.

Our comprehensive review may also yield permanent rule or policy changes that will apply to the proposed PFS facility and to other NRC-related facilities. The review process is ongoing and cumulative. It has already resulted in a number of security-related actions to address terrorism threats at both active and defunct nuclear facilities.

56 NRC at 343. Among these actions is the establishment of an Office of Nuclear Security and Incident Response, “responsi*1031ble for immediate operational security and safeguards issues as well as for long-term policy development!]] working] closely with law enforcement agencies and the Office of Homeland Security!]] ... coordinating] the NRC’s ongoing comprehensive security review.” Id. at 344-45.

We find it difficult to reconcile the Commission’s conclusion that, as a matter of law, the possibility of a terrorist attack on a nuclear facility is “remote and speculative,” with its stated efforts to undertake a “top to bottom” security review against this same threat. Under the NRC’s own formulation of the rule of reasonableness, it is required to make determinations that are consistent with its policy statements and procedures. Here, it appears as though the NRC is attempting, as a matter of policy, to insist on its preparedness and the seriousness with which it is responding to the post-September 11th terrorist threat, while concluding, as a matter of law, that all terrorist threats are “remote and highly speculative” for NEPA purposes.8

In sum, in considering the policy goals of NEPA and the rule of reasonableness that governs its application, the possibility of terrorist attack is not so “remote and highly speculative” as to be beyond NEPA’s requirements.

B

The NRC’s reliance upon the second PFS factor, that the Risk of a Terrorist Attack Cannot be Adequately Determined, 56 NRC at 350, is also not reasonable. First, the NRC’s dismissal of the risk of terrorist attacks as “unquantifiable” misses the point. The numeric probability of a specific attack is not required in order to assess likely modes of attack, weapons, and vulnerabilities of a facility, and the possible impact of each of these on the physical environment, including the assessment of various release scenarios. Indeed, this is precisely what the NRC already analyzes in different contexts. It is therefore possible to conduct a low probability-high consequence analysis without quantifying the precise probability of risk. The NRC itself has recognized that consideration of uncertain risks may take a form other than quantitative “probabilistic” assessment. In its “Proposed Policy Statement on Severe Accidents and Related Views on Nuclear Reactor Regulation,” 48 Fed. Reg. 16,014 (1983), the Commission stated that:

In addressing potential accident initiators (including earthquakes, sabotage, and multiple human errors) where empirical data are limited and residual uncertainty is large, the use of conceptual modeling and scenario assumptions in Safety Analysis Reports will be helpful. They should be based on the best qualified judgments of experts, either in the form of subjective numerical probability *1032estimates or qualitative assessments of initiating events and casual [sic] linkages in accident sequences.

48 Fed.Reg. at 16,020 (emphasis added).

No provision of NEPA, or any other authority cited by the Commission, allows the NRC to eliminate a possible environmental consequence from analysis by labeling the risk as “unquantifiable.” See Limerick Ecology Action, Inc. v. NRC, 869 F.2d 719, 754 (3rd Cir.1989) (J. Scirica, dissenting) (finding no “statutory provision, no NRC regulation or policy statement, and no case law that permits the NRC to ignore any risk found to be unquantifiable”). If the risk of a terrorist attack is not insignificant, then NEPA obligates the NRC to take a “hard look” at the environmental consequences of that risk. The NRC’s actions in other contexts reveal that the agency does not view the risk of terrorist attacks to be insignificant. Precise quantification is therefore beside the point.

Even if we accept the agency’s argument, the agency fails to adequately show that the risk of a terrorist act is unquantifiable. The agency merely offers the following analysis as to the quantifiability of a potential terrorist attack:

The horrors of September 11 notwithstanding, it remains true that the likelihood of a terrorist attack being directed at a particular nuclear facility is not quantifiable. Any attempt at quantification or even qualitative assessment would be highly speculative. In fact, the likelihood of attack cannot be ascertained with confidence by any state-of-the-art methodology. That being the case, we have no means to assess, usefully, the risks of terrorism at the PFS facility.

56 NRC at 350. The agency nonetheless has simultaneously shown the ability to conduct a “top to bottom” terrorism review. This leaves the Commission in the tenuous position of insisting on the impossibility of a meaningful, i.e. quantifiable, assessment of terrorist attacks, while claiming to have undertaken precisely such an assessment in other contexts. Further, as we have noted, the NRC has required site-specific analysis of such threats, involving numerous recognized scenarios.9

Thus, we conclude that precise quantification of a risk is not necessary to trigger NEPA’s requirements, and even if it were, the NRC has not established that the risk of a terrorist attack is unquantifiable.

C

The NRC’s third ground, that it is not required to conduct a “worst-case” analysis, is a non sequitur. Although it is a true statement of the law, the agency errs in equating an assessment of the environmental impact of terrorist attack with a demand for a worst-case analysis.

The Council on Environmental Quality (“CEQ”) regulations, 40 C.F.R. §§ 1500.1-1518.4, promulgated with the “purpose [of] tellfing] federal agencies what they must do to comply with [NEPA] procedures and achieve the goals of [NEPA],” have been interpreted by the Supreme Court as “entitled to substantial deference.” Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 355, 109 S.Ct. 1835, 104 L.Ed.2d 351 (citing Andrus v. Sierra Club, 442 U.S. 347, 358, 99 S.Ct. 2335, 60 *1033L.Ed.2d 943 (1979)). These regulations mandated worst-case analyses until 1986, when CEQ replaced the former 40 C.F.R. § 1502.22, requiring an agency, when relevant information was either unavailable or too costly to obtain, to include in the EIS a “worst-case analysis and an indication of the probability or improbability of its occurrence,” with the new and current version of the regulation, which requires an agency to instead deal with uncertainties by including within the EIS “a summary of existing credible scientific evidence which is relevant to evaluating the reasonable foreseeable significant adverse impacts on the human environment, and ... the agency’s evaluation of such impacts based upon theoretical approaches or research methods generally accepted in the scientific community.” 40 C.F.R. §§ 1502.22(b)(3), (4). The current requirement applies to those events with potentially catastrophic consequences “even if their probability of occurrence is low, provided that the analysis of impacts is supported by credible scientific evidence, is not based on pure conjecture, and is within the rule of reason.” 40 C.F.R. § 1502.22(b)(4). The Supreme Court held in Robertson that the amendment of the regulations had nullified the worst-case analysis requirement. 490 U.S. at 355, 109 S.Ct. 1835; Edwardsen v. U.S. Dep’t of Interior, 268 F.3d 781, 785 (9th Cir.2001).

The Commission is therefore correct when it argues that NEPA does not require a worst-case analysis. It is mistaken, however, when it claims that “Petitioners’ request for an analysis of [the environmental effects of] a successful terrorist attack at the Diablo Canyon ISFSI approximates a request for a ‘worst-case’ analysis that has long since been discarded by the CEQ regulations ... and discredited by the Federal courts.” According to the NRC, “Making the various assumptions required by [Petitioners’ scenario requires the NRC to venture into the realm of ‘pure conjecture.’ ” We disagree.

An indication of what CEQ envisioned when it imposed the worst-case analysis requirement can be gleaned from a 1981 CEQ memorandum, Forty Most Asked Questions Concerning CEQ’s National Environmental Policy Act Regulations, reprinted at 46 FR 18026-01 (March 23, 1981). CEQ answered one of those questions, “[w]hat is the purpose of a worst-case analysis? How is it formulated and what is the scope of the analysis?” with the following:

The purpose of the analysis is to ... cause agencies to consider th[ ]e potential consequences[of agency decisions] when acting on the basis of scientific uncertainties or gaps in available information. The analysis is formulated on the basis of available information, using reasonable projections of the worst possible consequences of a proposed action.

For example, if there are scientific uncertainty and gaps in the available information concerning the numbers of juvenile fish that would be entrained in a cooling water facility, the responsible agency must disclose and consider the possibility of the loss of the commercial or sport fishery. In addition to an analysis of a low probability/catastrophic impact event, the worst-case analysis should also include a spectrum of events of higher probability but less drastic impact.

46 FR 18026, 18032. While it is true that the agency is not required to consider consequences that are “speculative,”10 the *1034NRC’s argument wrongly labels a terrorist attack the worst-case scenario because of the low or indeterminate probability of such an attack. The CEQ memo, by including as worst-case scenarios events of both higher and lower probability, reveals that worst-case analysis is not defined solely by the low probability of the occurrence of the events analyzed, but also by the range of outcomes of those events. See also Greater Yellowstone Coalition v. Flowers, 321 F.3d 1250, 1260 (10th Cir.2003) (citing a witness’s testimony that the loss of bald eagle nesting sites was both “likely” and “a worst-case scenario”). Petitioners do not seek to require the NRC to analyze the most extreme (i.e., the “worst”) possible environmental impacts of a terrorist attack. Instead, they seek an analysis of the range of environmental impacts likely to result in the event of a terrorist attack on the Storage Installation. We reject the Commission’s characterization of this request as a demand for a worst-case analysis.

D

The NRC’s reliance on the fourth PFS factor, that it cannot comply with its NEPA mandate because of security risks, is also unreasonable. There is no support for the use of security concerns as an excuse from NEPA’s requirements. While it is true, as the agency claims, that NEPA’s requirements are not absolute, and are to be implemented consistent with other programs and requirements, this has never been interpreted by the Supreme Court as excusing NEPA’s application to a particularly sensitive issue. See Weinberger v. Catholic Action of Hawaii, 454 U.S. 139, 102 S.Ct. 197, 70 L.Ed.2d 298 (1981) (holding that the Navy was required to perform a NEPA review and to factor its results into decisionmaking even where the sensitivity of the information involved meant that the NEPA results could not be publicized or adjudicated). Weinberger can support only the proposition that security considerations may permit or require modification of some of the NEPA procedures, not the Commission’s argument that sensitive security issues result in some kind of NEPA waiver.

The application of NEPA’s requirements, under the rule of reason relied on by the NRC, is to be considered in light of the two purposes of the statute: first, ensuring that the agency will have and will consider detailed information concerning significant environmental impacts; and, second, ensuring that the public can both contribute to that body of information, and can access the information that is made public. Pub. Citizen, 541 U.S. at 768, 124 S.Ct. 2204. To the extent that, as the NRC argues, certain information cannot be publicized, as in Weinberger, other statutory purposes continue to mandate NEPA’s application. For example, that the public cannot access the resulting information does not explain the NRC’s determination to prevent the public from contributing information to the decision-making process. The NRC simply does not explain its unwillingness to hear and consider the information that Petitioners seek to contribute to the process, which would fulfill both the information-gathering and the public participation functions of NEPA. These arguments explain why a Weinberger-style limited proceeding might be appropriate, but cannot support the *1035NRC’s conclusion that NEPA does not apply. As we stated in No GWEN: “There is no ‘national defense’ exception to NEPA ... ‘The Navy, just like any federal agency, must carry out its NEPA mandate to the fullest extent possible and this mandate includes weighing the environmental costs of the [project] even though the project has serious security implications.’” 855 F.2d at 1384 (quoting Concerned About Trident v. Rumsfeld, 555 F.2d 817, 823 (D.C.Cir.1977)).

E

In sum, none of the four factors upon which the NRC relies to eschew consideration of the environmental effects of a terrorist attack satisfies the standard of reasonableness. We must therefore grant the petition in part and remand for the agency to fulfill its responsibilities under NEPA.

Our identification of the inadequacies in the agency’s NEPA analysis should not be construed as constraining the NRC’s consideration of the merits on remand, or circumscribing the procedures that the NRC must employ in conducting its analysis. There remain open to the agency a wide variety of actions it may take on remand, consistent with its statutory and regulatory requirements. We do not prejudge those alternatives. Nor do we prejudge the merits of the inquiry. We hold only that the NRC’s stated reasons for categorically refusing to consider the possibility of terrorist attacks cannot withstand appellate review based on the record before us.

We are also mindful that the issues raised by the petition may involve questions of national security, requiring sensitive treatment on remand. However, the NRC has dealt with our nation’s most sensitive nuclear secrets for many decades, and is well-suited to analyze the questions raised by the petition in an appropriate manner consistent with national security.

VI

We deny the petition as to the claims under the AEA and the APA. However, because we conclude that the NRC’s determination that NEPA does not require a consideration of the environmental impact of terrorist attacks does not satisfy reasonableness review, we hold that the EA prepared in reliance on that determination is inadequate and fails to comply with NEPA’s mandate. We grant the petition as to that issue and remand for further proceedings consistent with this opinion.

PETITION GRANTED IN PART; DENIED IN PART; REMANDED.

OPINION

ROTH, Circuit Judge:

The issue presented by this appeal is whether the Nuclear Regulatory Commission (NRC), when it is reviewing an application to relicense a nuclear power facility, must examine the environmental impact of a hypothetical terrorist attack on that nuclear power facility. The New Jersey Department of Environmental Protection (NJDEP) contends that the National Environmental Policy Act of 1969 (NEPA), 42 U.S.C. § 4321 et seq, requires the analysis of the impact of such an attack. NJDEP has petitioned for review of an NRC decision denying its request to intervene in relicensing proceedings for the Oyster Creek Nuclear Generating Station (Oyster Creek). The NRC concluded that terrorist attacks are “too far removed from the natural or expected consequences of agency action” to require an environmental impact analysis and that, in any event, it had already addressed the environmental impact of a potential terrorist act at Oyster Creek through its Generic Environmental Impact Statement and site-specific Supplemental Environmental Impact Statement. We agree with the NRC and will deny the petition.

I. BACKGROUND

A. Statutory and Regulatory Framework

The Atomic Energy Act of 1954(AEA), as amended, 42 U.S.C. § 2011 et seq., establishes a “comprehensive regulatory framework for the ongoing review of nuclear power plants located in the United States.” Sections 103 and 104(b) of the AEA authorize the NRC to issue licenses to operate commercial power reactors. 42 U.S.C. §§ 2133, 2134(b). Section 103 limits licenses to forty-year terms but provides for renewal of nearly-expired licenses. 42 U.S.C. § 2133. By regulation, the NRC may renew a license for up to twenty years. See 10 C.F.R. § 54.31.

Two sets of regulatory requirements govern the NRC’s review of license renewal applications. Under 10 C.F.R. Part 54, the NRC conducts a health and safety review focused on “the detrimental effects of aging” on the plant. See Nuclear Power Plant License Renewal: Revisions, 60 Fed.Reg. 22,461, 22,464 (May 8, 1995).

Under 10 C.F.R. Part 51, the NRC completes a NEPA-based environmental review, focusing on the potential impacts of twenty additional years of operation. NEPA is a procedural statute that does not mandate particular substantive results. Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 350-51, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989). Rather, it is designed “to insure a fully informed and well-considered decision” in the examination of potential environmental impacts of a proposed agency action. Vermont Yan *134 kee, 435 U.S. at 558, 98 S.Ct. 1197. NEPA “merely prohibits uninformed — rather than unwise — agency action.” Robertson, 490 U.S. at 351, 109 S.Ct. 1835. In addition, NEPA review should be consistent with NEPA’s “national policy [to] encourage productive and enjoyable harmony between man and his environment.” 42 U.S.C. § 4321. NEPA’s “twin aims” are to “ ‘placet ] upon an agency the obligation to consider every significant aspect of the environmental impact of a proposed action’ [and to] ensur[e] that the agency will inform the public that it has indeed considered environmental concerns in its decisionmaking process.” Baltimore Gas & Elec. Co. v. NRDC, 462 U.S. 87, 97, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983) (quoting Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 553, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978)).

By regulation, the NRC has divided the environmental requirements for license renewal into generic and plant-specific issues. This division resulted from “a systematic inquiry into the environmental impacts of refurbishment activities associated with license renewal and the environmental impacts of continued operation during the renewal period (up to 20 years for each licensing action).” Notice of Intent to Prepare an Environmental Impact Statement for the License Renewal of Nuclear Power Plants and to Conduct Scoping Process, 68 Fed.Reg. 332909, 33209 (June 3, 2003). The NRC analyzed “[t]he significance of environmental impacts ... for each of nearly 100 issues [and] categorized which of these analyses could be applied to all plants and whether the additional mitigation measures would be warranted for each environmental issue.” Id. Ultimately, “[o]f the 92 issues analyzed, 69 were resolved generally, 21 require a further site-specific analysis that applicants are required to address, and 2 require a site-specific assessment by the NRC.” Id.

The NRC’s “Generic Environmental Impact Statement for License Renewal of Nuclear Plants,” Final Report, Vol. I (May 1996) (GEIS), addresses issues that are common to all nuclear plants. These have been designated “Category 1” issues. GEIS at 1-5,1-6. Of particular note here, the GEIS reviews the risk of sabotage to nuclear power plants. The NRC has determined from this review that the risk is small and is provided for in the consideration of internal severe accidents:

The regulatory requirements under 10 CFR part 73 [i.e., “Physical Protection of Plants and Materials”] provide reasonable assurance that the risk from sabotage is small. Although the threat of sabotage events cannot be accurately quantified, the commission believes that acts of sabotage are not reasonably expected. Nonetheless, if such events were to occur, the commission would expect that resultant core damage and radiological releases would be no worse than those expected from internally initiated events.

Based on the above, the commission concludes that the risk from sabotage is small and additionally, that the risks f[ro]m other external events[] are adequately addressed by a generic consideration of internally initiated severe accidents.

GEIS at 5-18. The NRC expressly incorporated the GEIS’s findings related to internal severe accidents into the NRC’s environmental review regulations. See 10 C.F.R. Part 51 Subpt. A, App. B, Table B-1.

Environmental impacts not discussed in the GEIS are designated “Category 2” issues and must be addressed in an applicant’s environmental report. Id. § 51.53(c)(3)(ii). Ultimately, NRC staff *135prepares a site-specific Supplemental Environmental Impact Statement (SEIS) for each plant. Id. § 51.95(c). The SEIS includes evaluations of site-specific Category 2 issues — including a consideration of “severe accident mitigation alternatives” (SAMAs) for those issues that have not previously been considered — and “new and significant information” regarding Category 1 issues.

As a part of the relicensing review process, NRC regulations permit anyone with an “interest” in a licensing proceeding to obtain a hearing on admissible safety and environmental “contentions.” See 10 C.F.R. § 2.309(a), (d). Such a person must file a petition to intervene demonstrating standing and that “the issue raised ... is within the scope of the proceeding.” Id. § 2.309(f)(l)(iii). Unless a party obtains a waiver from the NRC, regulations are not “subject to attack” during adjudications. Id. § 2.335(a).

B. Factual and Procedural Background

On July 22, 2005, the AmerGen Energy Company, LLC (AmerGen) applied to the NRC to renew its operating license at Oyster Creek for an additional twenty years. Oyster Creek is located adjacent to Barnegat Bay in Lacey and Ocean Townships, Ocean County, New Jersey. Oyster Creek’s current license expires in April 2009. On September 15, 2005, the NRC published a notice of opportunity for hearing in the Federal Register. See Notice of Opportunity for Hearing Regarding Renewal of Facility Operating License No. DRP-16 for an Additional 20-Year Period, 70 Fed.Reg. 54,585 (Sept. 15, 2005).

On November 14, 2005, NJDEP filed a petition to intervene raising three contentions, only one of which it has raised in the appeal before us.1 NJDEP challenges the NRC’s failure to prepare an environmental impact statement (EIS) to study the effects of an aircraft attack on Oyster Creek. NJDEP contends that such an EIS should have contained, within its SAMAs analysis, a design basis threat (DBT) analysis2 and an analysis of mitigation alternatives for core melt sequences likely to result from an aircraft attack. The claims were reviewed by the Atomic Safety and Licensing Board (Board), which “held that terrorism and ‘design basis threat’ reviews, while important and ongoing, lie outside the scope of NEPA in general and of license renewal in particular.” See In re Amergen Energy Co., 65 N.R.C. 124, 128 (2007).

NJDEP appealed this decision to the NRC, which denied the claim. Id. at 126. The NRC agreed with the Board that terrorism concerns are security issues, which are not addressed during license renewal because they do not relate to the aging of the facility. Id. The NRC also found that NEPA “ ‘imposes no legal duty on the NRC to consider intentional malevolent acts’ ” because such acts are “ ‘too far removed from the natural or expected consequences of agency action.’ ” Id. at 129 (quoting the Board decision). Finally, the NRC found that a terrorism review would be redundant because (1) “the NRC has undertaken extensive efforts to enhance security at nuclear facili*136ties,” which it characterized as the best mechanism to protect the public; id. at 130; (2) the GEIS had concluded that “the core damage and radiological release from [terrorist] acts would be no worse than the damage and release to be expected from internally initiated events”; id. at 131; and (3) in its SEIS for Oyster Creek, the NRC had performed a site-specific SAMAs assessment; id. at 132.3

NJDEP filed a petition for review of the NRC’s order. We have jurisdiction pursuant to 28 U.S.C. § 2342(4).

II. DISCUSSION

NJDEP’s petition suffers from two insurmountable flaws, each of which independently supports our denial.4 First, NJDEP has not shown that there is a “reasonably close causal relationship” between the Oyster Creek relicensing proceeding and the environmental effects of a hypothetical aircraft attack. Accordingly, such an attack does not warrant NEPA evaluation. See DOT v. Pub. Citizen, 541 U.S. 752, 767, 124 S.Ct. 2204, 159 L.Ed.2d 60 (2004); Metro. Edison Co. v. People Against Nuclear Energy, 460 U.S. 766, 774, 103 S.Ct. 1556, 75 L.Ed.2d 534 (1983). Second, the NRC has already considered the environmental effects of a hypothetical terrorist attack on a nuclear plant and found that these effects would be no worse than those caused by a severe accident. NJDEP has not provided any evidence to challenge this conclusion and has not demonstrated that the NRC could undertake a *137more meaningful analysis of the specific risks associated with an aircraft attack on Oyster Creek. See Limerick, 869 F.2d at 744 & n. 31.

A. Causation

In rejecting NJDEP’s contention, the NRC held that “there simply is no proximate cause link between an NRC licensing action, such as [in this case] renewing an operating license, and any altered risk of terrorist attack. Instead, the level of risk depends upon political, social, and economic factors external to the NRC licensing process.” See In re AmerGen Energy Co., 65 N.R.C. at 130. NJDEP, on the other hand, asserts that the government has a duty to protect against foreseeable danger, even if that danger comes from intentional criminal conduct, and that here the risk of environmental harm caused by terrorists is foreseeable given the September 11, 2001, attacks on the World Trade Center and Oyster Creek’s proximity to important urban centers.5 NJDEP also finds significant the NRC’s efforts to improve security at nuclear facilities, asserting that these efforts demonstrate the NRC’s recognition that a terrorist attack is foreseeable.

The Supreme Court has spoken on two occasions regarding the circumstances in which NEPA requires an agency to prepare an EIS. The first concerned the resumption of activity at the Three Mile Island nuclear power plant after a serious accident caused a shutdown of one of the reactors. See Metro. Edison Co., 460 U.S. at 768, 103 S.Ct. 1556. Though no radiation was released in the accident, it caused widespread concern about the safety of the plant. Id. at 769, 103 S.Ct. 1556. A group of Harrisburg residents, organized as People Against Nuclear Energy (PANE), argued that restarting the reactor would “cause both severe psychological health damage to persons living in the vicinity[ ] and serious damage to the stability, cohesiveness, and well-being of the neighborhood communities.” Id. The NRC declined to take evidence on this issue, and PANE petitioned for review, arguing that both NEPA and the AEA required such an analysis. Id. at 770, 103 S.Ct. 1556. The D.C. Circuit agreed as to NEPA, finding, “NEPA requires agencies to consider effects on health. An effect on psychological health is an effect on health. Therefore, NEPA requires agencies to consider the effects on psychological health....” Id. at 771, 103 S.Ct. 1556.

The Supreme Court reversed.6 First, the Court noted that “NEPA does not require the agency to assess every impact or effect of its proposed action, but only the impact or effect on the environment.” Id. at 772, 103 S.Ct. 1556. The Court held that, in order to determine when NEPA requires consideration of a particular environmental effect, agencies and reviewing courts “must look at the relationship between that effect and the change in the physical environment caused by the major federal action at issue.” Id. at 773, 103 S.Ct. 1556. The Court then explained that NEPA attaches only when there is a “reasonably close causal relationship between a change in the physical environment and the effect at issue.” The Court likened *138this relationship to “the familiar doctrine of proximate cause from tort law.” Id. at 774, 103 S.Ct. 1556. In applying this standard to the case before it, the Court observed that the renewed operation of the reactor would affect the environment, particularly in the release of low levels of radiation, increased fog, the release of warm water into the Susquehanna River, and the potential results of a nuclear accident.7 Id. at 775, 103 S.Ct. 1556. It then observed that the NRC had considered all of these effects. The Court, however, found damage to psychological health caused by the perception of a risk of a nuclear accident too attenuated: “In a causal chain from renewed operation ... to psychological health damage, the element of risk and its perception by PANE’S members are necessary middle links. We believe that the element of risk lengthens the causal chain beyond the reach of NEPA.” Id.

The Supreme Court again discussed NEPA’s causation requirement in Department of Transportation v. Public Citizen, 541 U.S. 752, 124 S.Ct. 2204, 159 L.Ed.2d 60 (2004). Public Citizen concerned the operation of Mexican tractor-trailer trucks in the United States. Prior to 1982, these trucks were certified to operate in the United States by the Interstate Commerce Commission. In 1982, Congress suspended this certification procedure in light of concerns about Mexico’s discriminatory treatment of American trucks operating in Mexico. Id. at 759, 124 S.Ct. 2204. The United States agreed, however, as part of the North American Free Trade Agreement (NAFTA) to phase out the moratorium. Id.

In 1994, the President lifted the moratorium but called for new regulations related to the certification of Mexican trucks seeking to operate in the United States. Accordingly, the Federal Motor Carrier Safety Administration (FMCSA), a division of the Department of Transportation, published proposed safety regulations and procedures for the certification of Mexican trucks. The FMCSA also prepared an environmental assessment (EA) focusing on the effects of its proposed regulations. Id. at 760-62, 124 S.Ct. 2204. The EA did not consider the environmental impact of increased Mexican truck traffic because the FMCSA attributed this increase not to the regulations but to NAFTA and the President’s decision to lift the moratorium. Id. at 761, 124 S.Ct. 2204. A citizen group petitioned for review, arguing that NEPA required such an analysis. Id. at 766, 124 S.Ct. 2204.

The Supreme Court upheld the FMCSA’s decision. The Court noted that an EIS is required only for “ ‘major Federal actions,’ ” defined to include “ ‘actions with effects that may be major and which are potentially subject to Federal control and responsibility.’ ” Id. at 763, 124 S.Ct. 2204 (quoting 40 C.F.R. § 1508.18). The Court then noted that “effects” were limited by regulation to (1) “[djireet effects, which are caused by the action and occur at the same time and place,” and (2) “indirect effects, which are caused by the action and are later in time or farther removed in distance, but are still reasonably foresee*139able.” Id. at 764, 124 S.Ct. 2204 (internal quotation marks and citation omitted).

The Court concluded that the increase in Mexican truck traffic was not an effect of the FMCSA’s action. First, the Court noted that the FMCSA does not have the authority to exclude Mexican trucks from the United States. Rather, pursuant to congressional mandate, the FMCSA must certify every truck that can meet the FMCSA’s regulations. Id. at 766, 124 S.Ct. 2204. Next, the Court considered the causal relationship between the agency action and the environmental impact, as required by Metropolitan Edison. The Court characterized the causation at issue as “ ‘but for’ causation, where an agency’s action is considered a cause of an environmental effect even when the agency has no authority to prevent the effect.” It declared that this form of “but for” causation is “insufficient to make an agency responsible for a particular effect under NEPA.” Id. at 767, 103 S.Ct. 1556.

The Public Citizen Court also rejected the petitioner’s argument under the rule of reason, stating that agencies need not prepare an EIS when it would serve “no purpose” under NEPA. Id. It noted NEPA’s twin aims: (1) to force agencies to consider environmental impact as part of its decision making, and (2) to make information available to the public so that it can play a role in the decision making process. Because the FMCSA cannot prevent the entry of Mexican trucks, an EIS addressing increased traffic would not affect its decision making. Id. at 768, 124 S.Ct. 2204. Moreover, the public information purpose would not be served since FMCSA could not react to the input received from the public. Id. at 768-69, 124 S.Ct. 2204. Accordingly, the Court agreed with the FMCSA that “the legally relevant cause of the entry of the Mexican trucks is not FMCSA’s action, but instead the actions of the President in lifting the moratorium and those of Congress in granting the President this authority.” Id. at 769, 124 S.Ct. 2204.

NJDEP argues that neither Metropolitan Edison nor Public Citizen is apposite, asserting that those decisions involved cause and effect relationships that are far more attenuated than the one presented here. We disagree. The Supreme Court has directed that we “draw a manageable line between those causal changes that may make an actor responsible for an effect and those that do not.” Id. at 767, 124 S.Ct. 2204 (quoting Metro. Edison, 460 U.S. at 774 n. 7, 103 S.Ct. 1556). In the cases, this line appears to approximate the limits of an agency’s area of control. For example, in Metropolitan Edison, the NRC could control the nuclear facility and its operation but not how individuals perceived the risks of renewed operation and the possibility of another accident; therefore, these risks were too remote to require a NEPA analysis. Likewise, in Public Citizen, the FMCSA controlled the certification process, but it could not control the admission or volume of Mexican trucks; the FMCSA’s role was limited to certification.

In the instant case, the NRC controls whether equipment within a facility is suitable for continued operation or could withstand an accident, but it has no authority over the airspace above its facilities, which is largely controlled by Congress and the Federal Aviation Administration (FAA). The NRC has explicitly noted its limited ability to address airborne threats, articulating its consistent view that “security from terrorist attacks on nuclear facilities [i]s best approached by enhancing aviation security, including intelligence gathering and security at airports and on airplanes.” Riverkeeper, Inc. v. Collins, 359 F.3d 156, 161 (2d Cir.2004); cf. Glass Packaging In *140 stitute v. Regan, 737 F.2d 1083, 1092 (D.C.Cir.1984) (“NEPA is meant to supplement federal agencies’ other nonenviron-mental objectives, not to transplant specific regulatory burdens from those expert agencies otherwise authorized to redress specific nonenvironmental problems and pointlessly to reimpose those objectives on other unqualified agencies.”). This view is shared by other federal agencies. See Richard A. Meserve, Statement Submitted by the Nuclear Regulatory Commission to the Subcomm. on Oversight and Investigations of the H. Comm, on Energy and Commerce 5 (2003) (noting that when there were reported threats to the airspace above nuclear facilities, the FAA and the Department of Defense, rather than the NRC, responded to protect the airspace).

NRC’s lack of control over airspace supports our holding that a terrorist aircraft attack lengthens the causal chain beyond the “reasonably close causal relationship” required by those cases. Indeed, an aircraft attack on Oyster Creek requires at least two intervening events: (1) the act of a third-party criminal and (2) the failure of all government agencies specifically charged with preventing terrorist attacks. We conclude that this causation chain is too attenuated to require NEPA review. Moreover, this conclusion is supported by traditional tort law concepts of causation.8

According to the Restatement (Second) of Torts, the criminal conduct of a third person is not a superseding cause of harm unless the original actor “realized or should have realized the likelihood that [an opportunity for a third person to commit a crime] might be created, and that a third person might avail himself of the opportunity.” Restatement (Second) of Torts § 448. The comments to the section clarify the circumstances in which an actor should anticipate third-party criminal conduct: (1) situations that “afford[ ] temptations to which a recognizable percentage of humanity is likely to yield” and (2) situations “created at a place where persons of peculiarly vicious type are likely to be” who might yield to the temptation, even though the average individual would not do so. Id § 448 cmt. b. NJDEP has not demonstrated that either condition is present here.

The Restatement also clarifies when “an intervening force is a superseding cause.” See id. § 442. Section 442 lists six factors for consideration: (1) whether the third party causes harm “different in kind from that which would otherwise have resulted from the actor’s negligence,” (2) whether the event appears extraordinary in light of circumstances at the time, (3) whether the intervening force operates “independently of any situation created by the actor’s negligence,” (4) whether the intervening act is “due to a third person’s act,” (5) whether the third person’s act is wrongful and would subject him to liability, and (6) the “degree of culpability of [the] wrongful act by [the] third party.” Id. These factors counsel against finding the NRC’s relicensing of Oyster Creek to be the proximate cause of environmental harm in a terrorist attack. The first factor cuts against the NRC because the consequences of a successful terrorist attack would be similar to the possible consequences of a severe accident. The remaining five factors, however, are in the NRC’s favor. Such an attack would certainly be *141“extraordinary,” as there has never been an airborne attack on a nuclear facility, any terrorist would be operating independently of the NRC, the intervening force would be due to a third-party terrorist, a terrorist attack is wrongful, and the degree of culpability of the terrorist would far exceed that of the NRC.

Our decision in Port Authority of New York & New Jersey v. Arcadian Corp., 189 F.3d 305 (3d Cir.1999), although decided under state law, further supports our conclusion. Port Authority arose in the wake of the 1993 World Trade Center bombing. Id. at 309. The plaintiffs alleged that the defendant fertilizer manufacturers were negligent in the manufacture and sale of the fertilizer used in the attack. Id. at 310. We held “as a matter of law that the World Trade Center bombing was not a natural or probable consequence of any design defect in defendants’ products. In addition, the terrorists’ actions were superseding and intervening events breaking the chain of causation.” Id. at 319; see also Gaines-Tabb v. ICI Explosives, USA, Inc., 160 F.3d 613, 618 (10th Cir.1998) (after Oklahoma City bombing, defendant fertilizer manufacturer held not responsible for the criminal conduct of bomber in using the fertilizer to make the bomb). Similarly, here, a terrorist attack on a nuclear facility would be a superseding cause of the environmental effects felt after an attack.

The government agencies specifically charged with preventing an airborne terrorist attack would also serve as intervening forces. As noted above, the NRC’s sphere of authority is limited to the facilities themselves and the equipment within them. A terrorist attack on an NRC-licensed facility would require, at a minimum, a failure by the FAA and the Department of Defense to protect and defend the facility.

An additional factor counsels against finding that the NRC’s relicensing of the Oyster Creek facility would be the proximate cause of environmental harm in the event of an airborne attack. In insisting that we “draw a manageable line” when imposing NEPA responsibilities, the Metropolitan Edison Court noted the limited time and resources of federal agencies and warned that “[t]he scope of the agency’s inquiries must remain manageable if NEPA’s goal of [ensuring] a fully informed and well considered decision is to be accomplished.” Id. at 774 n. 7, 776, 103 S.Ct. 1556 (internal quotation marks omitted). Applied to the case before it, the Court indicated that, if agencies were required to assess psychological health damage associated with increased risk, agencies would “expend considerable resources” on issues “not otherwise relevant to their congres-sionally assigned functions” and “resources may be spread so thin that agencies are unable adequately to pursue protection of the physical environment and natural resources.” Id.

Similarly, if NEPA required the NRC to analyze the potential consequences of an airborne attack, the NRC would spend time and resources assessing security risks over which it has little control and which would not likely aid its other assigned functions to assure the safety and security of nuclear facilities. Moreover, an analysis of the risks of a terrorist attack on Oyster Creek, as well as NJDEP’s arguments concerning Oyster Creek’s status as a particularly vulnerable terrorist target, implicate security concerns that are broader than those at issue under NEPA. For example, security decisions must be centralized rather than made on a site-specific basis since those in charge of each site may have differing ideas over how to spread limited resources. This policy is reflected in NRC regulations, which sepa*142rate its health and safety review, conducted through rulemaking under the APA, from the environmental review required by NEPA. See 10 C.F.R. Parts 51, 54; In re AmerGen Energy Co., 65 N.R.C. at 130. Likewise, security reviews involve analysis of sensitive information not available to the public, while NEPA requires public participation and transparency. See In re Private Fuel Storage, CLI-02-023, 56 N.R.C. 340 (Dec. 18, 2002).9

In holding that there is no “reasonably close causal relationship” between a reli-censing proceeding and the environmental effects of an aircraft attack on the licensed facility, we depart from the reasoning of the Ninth Circuit Court of Appeals in San Luis Obispo Mothers for Peace v. NRC, 449 F.3d 1016 (9th Cir.2006). The Mothers for Peace court held that, given “the policy goals of NEPA and the rule of reasonableness that governs its application, the possibility of terrorist attack is not so ‘remote and highly speculative’ as to be beyond NEPA’s requirements.” Id. at

1031. We note, initially, that Mothers for Peace is distinguishable on the ground that it involved the proposed construction of a new facility — a change to the physical environment arguably with a closer causal relationship to a potential terrorist attack than the mere relicensing of an existing facility. See id. at 1021. More centrally, however, we disagree with the rejection of the “reasonably close causal relationship” test set forth by the Supreme Court and hold that this standard remains the law in this Circuit.10 We also note that no other circuit has required a NEPA analysis of the environmental impact of a hypothetical terrorist attack. See Mid States Coalition for Progress v. Surface Transp. Bd., 345 F.3d 520, 544 (8th Cir.2003) (holding that agency did not err in declining to reopen record for construction of new rail lines in light of terrorist attacks of September 11, 2001); Limerick Ecology Action v. NRC, 869 F.2d 719, 743-44 (3d Cir.1989) (upholding NRC decision not to analyze risks of sabotage under NEPA where petitioner *143did not propose a meaningful way to analyze the risk); Glass Packaging Inst., 737 F.2d at 1091 (upholding agency decision not to consider possibility that a “deranged criminal” might tamper with bottles); City of New York v. Dep’t of Transp., 715 F.2d 732, 750 (1983) (deferring to agency’s conclusion that risks of sabotage “were too far afield for consideration” in the NEPA analysis of regulation governing highway shipment of radioactive material).

Finally, NJDEP’s argument concerning the relevance of the NRC’s other efforts to prevent terrorist attacks is misplaced. As the NRC notes, even the Ninth Circuit Court of Appeals has held that precautionary actions to guard against a particular risk do not trigger a duty to perform a NEPA analysis. See Ground Zero Ctr. for Non-Violent Action v. Dep’t of the Navy, 383 F.3d 1082, 1090-91 (9th Cir.2004) (fact that the Navy took potential Trident missile accident into account when planning base layout did not mean, in and of itself, that Navy had to prepare NEPA review outlining effects of that potential accident).

In sum, the NRC correctly concluded that the relicensing of Oyster Creek does not have a “reasonably close causal relationship” with the environmental effects that would be caused in the event of a terrorist attack.

B. The NRC’S Prior Analysis of the Terrorism Threat

Even if NEPA required an assessment of the environmental effects of a hypothetical terrorist attack on a nuclear facility, the NRC has already made this assessment. As described above, the GEIS addresses the risks associated with a terrorist attack, stating that “estimates of risk from sabotage” are impossible to quantify but nonetheless characterizing the risks as “small.” GEIS at 5-18. The GEIS goes on to say that, should the unlikely event occur, the effects would be “no worse than those expected from internally initiated events.” Id. The NRC rules codify these generic findings, and by regulation, license renewal applicants are excused from discussing generic issues in their environmental reports. See 10 C.F.R. § 51.53(c)(3)(i); id. Part 51 Subpt. A, App. B, Table B.

Generic analysis “is clearly an appropriate method of conducting the hard look required by NEPA.” Baltimore Gas, 462 U.S. at 101, 103 S.Ct. 2246 (internal quotation marks omitted). Indeed, it is “horn-book administrative law that an agency need not — indeed should not — entertain a challenge to a regulation” in an individual adjudication. Tribune Co. v. FCC, 133 F.3d 61, 68 (D.C.Cir.1998). NJDEP’s contention challenges the NRC’s generic findings, essentially arguing that certain characteristics of Oyster Creek make the risk of a terrorist attack more than “small” and the environmental effects of a terrorist attack somehow different from “those expected from internally initiated events.” These arguments thus amount to collateral attacks on the licensing renewal regulations, and the proper way to raise them would have been in a petition for rulemak-ing or a petition for a waiver based on “special circumstances.” See 10 C.F.R. §§ 2.335, 2.802.11

Moreover, the NRC prepared a SEIS that analyzed alternatives at Oyster Creek to mitigate severe accidents. See SEIS at *1445-3 through 5-12. Accordingly, the GEIS and SEIS together provide both generic and site-specific analyses of potential environmental impacts at Oyster Creek arising from terrorist attacks. New Jersey has never explained how or why an aircraft attack on Oyster Creek would produce impacts that are different from severe accidents and has not provided any evidence that the NRC could engage in a meaningful analysis of the risks of an attack. Instead, NJDEP argues, quoting our decision in Limerick Ecology Action v. NRC, that the NRC’s “mere assertion of unquan-tifiability” does not immunize it from having to conduct a NEPA analysis. See 869 F.2d at 744 n. 31. This is a true statement of the law, but it ignores our holding in Limerick that the burden is on the petitioner to demonstrate that the NRC could evaluate risks more meaningfully than it has already done. See id. at 744 n. 31. NJDEP has not met its burden here.

III. CONCLUSION

Because NJDEP did not present an admissible contention before the NRC, concerning the environmental effects of a hypothetical aircraft attack on Oyster Creek, we will deny the petition for review.

Mr. Chief Justice Burger

delivered the opinion of the Court.

These appeals present the- question of whether Congress may, consistent with the Constitution, impose a limitation on *63liability for nuclear accidents resulting from the operation of private nuclear power plants licensed by the Federal Government.

I

A

When Congress passed the Atomic Energy Act of 1946, it contemplated that the development of nuclear power would be a Government monopoly. See Act of Aug. 1, 1946, ch. 724, 60 Stat. 755. Within a decade, however, Congress concluded that the national interest would be best served if the Government encouraged the private sector to become involved in the development of atomic energy for peaceful purposes under a program of federal regulation and licensing. See H. R. Rep. No. 2181, 83d Cong., 2d Sess., 1-11 (1954). The Atomic Energy Act of 1954, Act of Aug. 30, 1954, ch. 1073, 68 Stat. 919, as amended, 42 U. S. C. §§ 2011-2281 (1970 ed. and Supp. V), implemented this policy decision, providing for licensing of private construction, ownership, and operation of commercial nuclear power reactors for energy production under strict supervision by the Atomic Energy Commission (AEC).1 See Power Reactor Development Co. v. Electrical Workers, 367 U. S. 396 (1961), rev’g and remanding 108 U. S. App. D. C. 97, 280 P. 2d 645 (I960).

Private industry responded to the Atomic Energy Act of 1954 with the development of an experimental power plant constructed under the auspices of a consortium of interested companies. It soon became apparent that profits from the private exploitation of atomic energy were uncertain and the accompanying risks substantial. See Green, Nuclear Power: *64Risk, Liability, and Indemnity, 71 Mich. L. Rev. 479-481 (1973) (Green). Although the AEC offered incentives to encourage investment, there remained in the path of the private nuclear power industry various problems — the risk of potentially vast liability in the event of a nuclear accident of a sizable magnitude being the major obstacle. Notwithstanding comprehensive testing and study, the uniqueness of this form of energy production made it impossible totally to rule out the risk of a major nuclear accident resulting in extensive damage. Private industry and the AEC were confident that such a disaster would not occur, but the very uniqueness of nuclear power meant that the possibility remained, and the potential liability dwarfed the ability of the industry and private insurance companies to absorb the risk. See Hearings before the Joint Committee on Atomic Energy on Government Indemnity for Private Licensees and AEC Contractors Against Reactor Hazards, 84th Cong., 2d Sess., 122-124 (1956). Thus, while repeatedly stressing that the risk of a major nuclear accident was extremely remote, spokesmen for the private sector informed Congress that they would be forced to withdraw from the field if their liability were not limited by appropriate legislation. Id., at 9, 109-110, 115, 120, 136-137, 148, 181, 195, and 240.

Congress responded in 1957 by passing the Price-Anderson Act, 71 Stat. 576, 42 U. S. C. § 2210 (1970 ed. and Supp. V). The Act had the dual purpose of “protect [ing] the public and . . . encouraging] the development of the atomic energy industry.” 42 U. S. C. § 2012 (i). In its original form, the Act limited the aggregate liability for a single nuclear incident2 to $500 million plus the amount of liability insurance *65available on the private market — some $60 million in 1957. The nuclear industry was required to purchase the maximum available amount of privately underwritten public liability insurance, and the Act provided that if damages from a nuclear disaster exceeded the amount of that private insurance coverage, the Federal Government would indemnify the licensee and other “persons indemnified” 3 in an amount not to exceed $500 million. Thus, the actual ceiling on liability was the amount of private insurance coverage plus the Government’s indemnification obligation which totaled $560 million.

Since its enactment, the Act has been twice amended, the first occasion being on the eve of its expiration in I960.4 These amendments extended the basic liability-limitation provisions for another 10 years, and added a provision which had the effect of requiring those indemnified under the Act to waive all legal defenses in the event of a substantial nuclear accident.5 This provision was based on a congressional concern that state tort law dealing with liability for nuclear incidents was generally unsettled and that some way of insuring a common standard of responsibility for all jurisdictions — strict liability — was needed. A waiver of defenses was thought to be the preferable approach since it entailed less *66interference with state tort law than would the enactment of a federal statute prescribing strict liability.6 See S. Rep. No. 1605, 89th Cong., 2d Sess., 6-10 (1966).

In 1975, Congress again extended the Act’s coverage until 1987, and continued the $560 million limitation on liability. However a new provision was added requiring, in the event of a nuclear incident, each of the 60 or more reactor owners to contribute between $2 and $5 million toward the cost of compensating victims.7 42 U. S. C. § 2210 (b) (1970 ed., Supp. V). Since the liability ceiling remained at the same level, the effect of the “deferred premium” provision was to reduce the Federal Government’s contribution to the liability pool.8 In its amendments to the Act in 1975, Congress also explicitly provided that “in the event of a nuclear incident involving damages in excess of [the] amount of aggregate liability, the Congress will thoroughly review the particular incident and will take whatever action is deemed necessary and appropriate to protect the public from the consequences of a *67disaster of such magnitude . . . 42 U. S. C. § 2210 (e) (1970 ed., Supp. Y).

Under the Price-Anderson Act as it presently stands, liability in the event of a nuclear incident causing damages of $560 million or more would be spread as follows: $315 million would be paid from contributions by the licensees of the 63 private operating nuclear power plants; $140 million would come from private insurance (the maximum now available); the remainder of $105 million would be borne by the Federal Government.9

B

Appellant in No. 77-262, Duke Power Co., is an investor-owned public utility which is constructing one nuclear power plant in North Carolina and one in South Carolina. Duke Power, along with the NRC, was sued by appellees, two organizations — Carolina Environmental Study Group and the Catawba Central Labor Union — and 40 individuals who live within close proximity to the planned facilities. The action was commenced in 1973, and sought, among other relief, a declaration that the Price-Anderson Act is unconstitutional.10

After the parties had engaged in extensive discovery, the District Court held an evidentiary hearing on the questions of whether the issues were ripe for adjudication and whether *68appellees had standing to challenge the constitutionality of the Act. That court determined that appellees had standing and that their claim could properly be adjudicated. The District Court went on to hold that the Price-Anderson Act was unconstitutional in two respects: (a) it violated the Due Process Clause of the Fifth Amendment because it allowed injuries to occur without assuring adequate compensation to the victims; (b) the Act offended the equal protection component of the Fifth Amendment by forcing the victims of nuclear incidents to bear the burden of injury, whereas society as a whole benefits from the existence and development of nuclear power.

We noted probable jurisdiction 11 in these appeals, 434 U. S. 937 (1977), and we now reverse.

II

As a threshold matter, we must address the question of whether the District Court had subject-matter jurisdiction over appellees’ claims, despite the fact that none of the parties raised this issue and the District Court did not consider it. See Liberty Mutual Ins. Co. v. Wetzel, 424 U. S. 737, 740 (1976). Appellees’ complaint alleges jurisdiction under 28 U. S. C. § 1337 (1976 ed.), which provides for original jurisdiction in the district courts over “any civil action or proceeding arising under any Act of Congress regulating commerce or protecting trade and commerce against restraints and monopolies.” Our reading of the pleadings,12 however, indicates that *69appellees’ claims do not “arise under” the Price-Anderson Act as that statutory language has been interpreted in prior decisions. See Peyton v. Railway Express Agency, 316 U. S. 350, 353 (1942).

Specifically, as we read the complaint, appellees are making two basic challenges to the Act — both of which find their moorings in the Fifth Amendment. First, appellees contend that the Due Process Clause protects them against arbitrary governmental action adversely affecting their property rights and that the Price-Anderson Act — -which both creates the source of the underlying injury and limits the recovery therefor — constitutes such arbitrary action. And second, they are contending that in the event of a nuclear accident their property would be “taken” without any assurance of just compensation.. The Price-Anderson Act is the instrument of the taking since on this record, without it, there would be no power plants and no possibility of an accident. Implicit in the complaint is also the assumption that there exists a cause of action directly under the Constitution to vindicate appellees’ federal rights through a suit against the NRC, the executive agency charged with enforcement and administration of the allegedly unconstitutional statute.13 Appellees’ right to relief *70thus depends not on the interpretation or construction of the Price-Anderson Act itself, but instead “upon the construction or application of the Constitution,” Smith v. Kansas City Title & Trust Co., 255 U. S. 180, 199 (1921). Hence, if there exists jurisdiction to hear appellees' claims at all, it must be derived from 28 U. S. C. § 1331 (a) (1976 ed.), the general federal-question statute, rather than from § 1337 — the jurisdictional base pleaded.14

For purposes of determining whether jurisdiction exists under § 1331 (a) to resolve appellees’ claims, it is not necessary to decide whether appellees’ alleged cause of action against the NRC based directly on the Constitution is in fact a cause of action “on which [appellees] could actually recover.” Bell v. Hood, 327 U. S. 678, 682 (1946). Instead, the test is whether “ ‘the cause of action alleged is so patently without merit as to justify . . . the court’s dismissal for want of jurisdiction.’ ” Hagans v. Lavine, 415 U. S. 528, 542-543 (1974), quoting Bell v. Hood, supra, at 683. (Emphasis added.) See also Oneida Indian Nation v. County of Oneida, 414 U. S. 661, 666 (1974) (test is whether right claimed is “so insubstantial, implausible, foreclosed by prior decisions of this *71Court, or otherwise completely devoid of merit as not to involve a federal controversy”)- In light of prior decisions, for example, Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388 (1971) and Hagans v. Lavine, supra, as well as the general admonition that “where federally protected rights have been invaded . . . courts will be alert to adjust their remedies so as to grant the necessary relief,” Bell v. Hood, supra, at 684, we conclude that appellees’ allegations are sufficient to sustain jurisdiction under § 1331 (a).15

The further question of whether appellees’ cause of action under the Constitution is one generally to be recognized need not be decided here. The question does not directly implicate our jurisdiction, see Bell v. Hood, supra, was not raised in the court below, was not briefed, and was not addressed during oral argument. As we noted last Term in a similar context, questions of this sort should not be resolved on such an inadequate record; leaving them unresolved is no bar to full consideration of the merits. See Mt. Healthy City Bd. of Educ. v. Doyle, 429 U. S. 274, 278-279 (1977). It is enough for present purposes that the claimed cause of action to vindicate appel-*72lees’ constitutional rights is sufficiently substantial and color-able to sustain jurisdiction under § 1331 (a).16

Ill

The District Judge held four days of hearings on the questions of standing and ripeness; his factual findings form the basis for our analysis of these issues.

A

The essence of the standing inquiry is whether the parties seeking to invoke the court’s jurisdiction have “alleged such a personal stake in the outcome of the controversy as to assure that concrete adverseness which sharpens the presentation of issues upon which the court so largely depends for illumination of difficult constitutional questions.” Baker v. Carr, 369 U. S. 186, 204 (1962). As refined by subsequent reformulation, this requirement of a “personal stake” has come to be understood to require not only a “distinct and palpable injury,” to the plaintiff, Warth v. Seldin, 422 U. S. 490, 501 (1975), but also a “fairly traceable” causal connection between the claimed injury and the challenged conduct. Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 261 (1977). See also Simon v. Eastern Ky. Welfare Rights Org., 426 U. S. 26, 41-42 (1976); Linda R. S. v. Richard D., 410 U. S. 614, 617 (1973). Application of these constitutional standards to the factual findings of the District Court persuades us that the Art. Ill requisites for standing are satisfied by appellees.

We turn first to consider the kinds of injuries the District Court found the appellees suffered. It discerned two categories of effects which resulted from the operation of nuclear *73power plants in potentially dangerous proximity to appellees’ living and working environment. The immediate effects included: (a) the production of small quantities of non-natural radiation which would invade the air and water; (b) a “sharp increase” in the temperature of two lakes presently used for recreational purposes resulting from the use of the lake waters to produce steam and to cool the reactor; (c) interference with the normal use of the waters of the Catawba River; (d) threatened reduction in property values of land neighboring the power plants; (e) “objectively reasonable” present fear and apprehension regarding the “effect of the increased radioactivity in air, land and water upon [appellees] and their property, and the genetic effects upon their descendants”; and (f) the continual threat of “an accident resulting in uncontrolled release of large or even small quantities of radioactive material” with no assurance of adequate compensation for the resultant damage. 431 F. Supp. 203, 209. Into a second category of potential effects were placed the damages “which may result from a core melt or other major accident in the operation of a reactor . . . .” Id., at 209.17

For purposes of the present inquiry, we need not determine whether all the putative injuries identified by the District Court, particularly those based on the possibility of a nuclear accident and the present apprehension generated by this future uncertainty, are sufficiently concrete to satisfy constitutional requirements. Compare O’Shea v. Littleton, 414 U. S. 488 (1974), with United States v. SCRAP, 412 U. S. 669 (1973). See also Conservation Society of Southern Vermont v. AEC, Civ. Action No. 19-72 (DC Apr. 17, 1975). It is enough that several of the “immediate” adverse effects were found to harm appellees. Certainly the environmental and aesthetic consequences of the thermal pollution of the two lakes in the vicinity of the disputed power plants is the type *74of harmful effect which has been deemed adequate in prior cases to satisfy the "injury in fact” standard. See United States v. SCRAP, supra. Cf. Sierra Club v. Morton, 405 U. S. 727, 734 (1972) ,18 And the emission of non-natural radiation into appellees’ environment would also seem a direct and present injury, given our generalized concern about exposure to radiation and the apprehension flowing from the uncertainty about the health and genetic consequences of even small emissions like those concededly emitted by nuclear power plants.19

The more difficult step in the standing inquiry is establishing that these injuries “fairly can be traced to the challenged action of the defendant,” Simon v. Eastern Ky. Welfare Rights Org., supra, at 41, or put otherwise, that the exercise of the Court’s remedial powers would redress the claimed injuries. 426 U. S., at 43. The District Court discerned a “but for” causal connection between the Price-Anderson Act, which appellees challenged as unconstitutional, “and the construction of the nuclear plants which the [appellees] view as a threat to them.” 431 F. Supp., at 219. Particularizing that causal link to the facts of the instant case, the District Court concluded that “there is a substantial likelihood that Duke would not be able to complete the construction and maintain the operation of the McGuire and Catawba Nuclear Plants *75but for the protection provided by the Price-Anderson Act.” Id., at 220.

These findings, which, if accepted, would likely satisfy the second prong of the constitutional test for standing as elaborated in Simon, 20 are challenged on two grounds. First, it is argued that the evidence presented at the hearing, contrary to the conclusion reached by the District Court, indicated that the McGuire and Catawba nuclear plants would be completed and operated without the Price-Anderson Act’s limitation on liability. And second, it is contended that the Price-Anderson Act is not, in some essential sense, the “but for” cause of the disputed nuclear power plants and resultant adverse effects since if the Act had not been passed Congress may well have chosen to pursue the nuclear program as a Government monopoly as it had from 1946 until 1954. We reject both of these arguments.

The District Court’s finding of a “substantial likelihood” that the McGuire and Catawba nuclear plants would be neither completed nor operated absent the Price-Anderson Act rested in major part on the testimony of corporate officials before the Joint Committee on Atomic Energy (JCAE) in 1956-1957 when the Price-Anderson Act was first considered and again in 1975 when a second renewal was discussed. During the 1956-1957 hearings, industry spokesmen for the utilities and the producers of the various component parts of the power plants expressed a categorical unwillingness to participate in the development of nuclear power absent guarantees of a limitation on their liability. 431 F. Supp., at 215. See also *76Green 486, 490-491.21 By 1975, the tenor of the testimony had changed only slightly. While large utilities and producers were somewhat more equivocal about whether a failure to renew Price-Anderson would entail their leaving the industry, the smaller producers of component parts and architects and engineers — all of whom are essential to the building of the reactors and generating plants — considered renewal of the Act as the critical variable in determining their continued involvement with nuclear power. 431 F. Supp., at 216-217. Duke Power itself, in its letter to the Committee urging extension of the Act, cited recent experiences with suppliers and contractors who were requiring the inclusion of cancellation clauses in their contracts to take effect if the liability-limitation provisions were eliminated. Id,., at 217. And the Report of the JCAE, in discussing the need for renewal of the Act, stated:

“Nuclear power plants now in the planning and design phases would not receive construction permits until about 1977-1978. Thus there is uncertainty as to whether these plants would receive protection in the form of Government indemnity. Reactor manufacturers and architect-engineers are already requiring escape clauses in their contracts to permit cancellation in the event some form of protection from unlimited potential liability is not provided. Action is required soon to prevent disruption in utility plans for nuclear power.” H. R. Rep. No. 94^-648, p. 7 (1975).

Nor was the testimony at the hearing in this case, evaluation of which is the primary responsibility of the trial judge, at odds with the impression drawn from the legislative history. The testimony of Executive Vice President Lee of Duke Power *77simply echoed the views presented by Duke and others to Congress in 1975, that is, although some of the utilities themselves might be confident enough with respect to safety factors to proceed with nuclear power absent a liability limitation, the suppliers of critical parts and the utility shareholders could reasonably be expected to take a more cautious view.22 Appellees presented expert testimony essentially to the same effect. Considering the documentary evidence and the testimony in the record, we cannot say we are left with “the definite and firm conviction that” the finding by the trial court of a substantial likelihood that the McGuire and Catawba nuclear power plants would be neither completed nor operated absent the Price-Anderson Act is clearly erroneous; and, hence, we are bound to accept it. United States v. United States Gypsum Co., 333 U. S. 364, 395 (1948).

The second attack on the District Court’s finding of a causal link warrants only brief attention. Essentially the argument is, as we understand it, that Price-Anderson is not a “but for” cause of the injuries appellees claim since, if Price-Anderson had not been passed, the Government would have undertaken development of nuclear power on its own and the same injuries would likely have accrued to appellees from such Government-operated plants as from privately operated ones. Whatever the ultimate accuracy of this speculation, it is not responsive to the simple proposition that private power companies now do in fact operate the nuclear-powered generating plants in*78juring appellees, and that their participation would not have occurred but for the enactment and implementation of the Prices Anderson Act.. Nothing in our prior cases requires a party seeking to invoke federal jurisdiction to negate the kind of speculative and hypothetical possibilities suggested in order to demonstrate the likely effectiveness of judicial relief.

B

It is further contended that in addition to proof of injury and of a causal link between such injury and the challenged conduct, appellees must demonstrate a connection between the injuries they claim and the constitutional rights being asserted. This nexus requirement is said to find its origin in Flast v. Cohen, 392 U. S. 83 (1968), where the general question of taxpayer standing was considered:

“The nexus demanded of federal taxpayers has two aspects to it. First, the taxpayer must establish a logical link between that status and the type of legislative enactment attacked. . . . Secondly, the taxpayer must establish a nexus between that status and the precise nature of the constitutional infringement alleged.” Id., at 102.

See also United States v. Richardson, 418 U. S. 166, 174 — 175 (1974). Since the environmental and health injuries claimed by appellees are not directly related to the constitutional attack on the Price-Anderson Act, such injuries, the argument continues, cannot supply a predicate for standing.23 We decline to accept this argument.

The major difficulty with the argument is that it implicitly assumes that the nexus requirement formulated in the context of taxpayer suits has general applicability in suits of all other types brought in the federal courts. No cases have been cited *79outside the context of taxpayer suits where we have demanded this type of subject-matter nexus between the right asserted and the injury alleged, and we are aware of none.24 Instead, in Schlesinger v. Reservists Comm, to Stop the War, 418 U. S. 208, 225 n. 15 (1974), we explicitly rejected such a broad compass for the Flast nexus requirement:

“Looking To the substantive issues' which Flast stated to be both ‘appropriate and necessary' in relation to taxpayer standing was for the express purpose of determining ‘whether there is a logical nexus between the [taxpayer] status asserted and the claim sought to be adjudicated.' 392 U. S., at 102. This step is not appropriate on a claim of citizen standing since the Flast nexus test is not applicable where the taxing and spending power is not challenged. . . .''

We continue to be of the same view and cannot accept the contention that, outside the context of taxpayers’ suits, a litigant must demonstrate something more than injury in fact and a substantial likelihood that the judicial relief requested will prevent or redress the claimed injury to satisfy the “case or controversy” requirement of Art. III.25

*80Our prior cases have, however, acknowledged “other limits on the class of persons who may invoke the courts’ decisional and remedial powers,” Warth v. Seldin, 422 U. S., at 499, which' derive from general prudential concerns “about the proper — and properly limited' — role of the courts in a democratic society.” Id., at 498. See also Schlesinger v. Reservists Comm. to Stop the War, supra, at 221-227. Thus, we have declined to grant standing where the harm asserted amounts only to a generalized grievance shared by a large number of citizens in a substantially equal measure. See United States v. Richardson, supra. We have also narrowly limited the circumstances in which one party will be given standing to assert the legal rights of another. “[E]ven when the plaintiff has alleged injury sufficient to meet the ‘case or controversy’ requirement, this Court has held that the plaintiff generally must assert his own legal rights and interests, and cannot rest his claim to relief on the legal rights or interests of third parties.” Warth v. Seldin, supra, at 499. See also United States v. Raines, 362 U. S. 17 (1960). This limitation on third-party standing arguably suggests a connection between the claimed injury and the right asserted bearing some resemblance to the nexus requirement now urged upon us.

There are good and sufficient reasons for this prudential limitation on standing when rights of third parties are implicated — the avoidance of the adjudication of rights which those not before the Court may not wish to assert, and the assurance that the most effective advocate of the rights at issue is present to champion them. See Singleton v. Wulff, 428 U. S. 106, 113-114 (1976). We do not, however, find these reasons a satisfactory predicate for applying this limitation or a similar nexus requirement to all cases as a matter of course. Where a party champions his own rights, and where the injury alleged is a concrete and particularized one which will be *81prevented or redressed by the relief requested, the basic practical and prudential concerns underlying the standing doctrine are generally satisfied when the constitutional requisites are met. See, e. g., Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252 (1977).

We conclude that appellees have standing to challenge the constitutionality of the Price-Anderson Act.26

C

The question of the ripeness of the constitutional challenges raised by appellees need not long detain us. To the extent that “issues of ripeness involve, at least in part, the existence of a live ‘Case or Controversy,’ ” Regional Rail Reorganization Act Cases, 419 U. S., at 138, our conclusion that appellees will sustain immediate injury from the operation of the disputed power plants and that such injury would be redressed by the relief requested would appear to satisfy this requirement.

The prudential considerations embodied in the ripeness doctrine also argue strongly for a prompt resolution oí the claims presented. Although it is true that no nuclear accident has yet occurred and that such an occurrence would eliminate much of the existing scientific uncertainty surrounding this *82subject, it would not, in our view, significantly advance our ability to deal with the legal issues presented nor aid us in their resolution. However, delayed resolution of these issues would foreclose any relief from the present injury suffered by appellees — relief that would be forthcoming if they were to prevail in their various challenges to the Act. Similarly, delayed resolution would frustrate one of the key purposes of the Price-Anderson Act — the elimination of doubts concerning the scope of private liability in the event of major nuclear accident. In short, all parties would be adversely affected by a decision to defer definitive resolution of the constitutional validity vel non of the Price-Anderson Act. Since we are persuaded that “we will be in no better position later than we are now” to decide this question, Id., at 143-145, we hold that it is presently ripe for adjudication.

IV

The District Court held that the Price-Anderson Act contravened the Due Process Clause because “[t]he amount of recovery is not rationally related to the potential losses”; because “[t]he Act tends to encourage irresponsibility in matters of safety and environmental protection . . .”; and finally because “[t]here is no quid pro quo” for the liability limitations. 431 F. Supp., at 222-223. An equal protection violation was also found because the Act “places the cost of [nuclear power] on an arbitrarily chosen segment of society, those injured by nuclear catastrophe.” Id., at 225. Application of the relevant constitutional principles forces the conclusion that these holdings of the District Court cannot be sustained.

A

Our due process analysis properly begins with a discussion of the appropriate standard of review. Appellants, portraying the liability-limitation provision as a legislative balancing of economic interests, urge that the Price-Anderson Act be *83accorded the traditional presumption of constitutionality generally accorded economic regulations and that it be upheld absent proof of arbitrariness or irrationality on the part of Congress. See Ferguson v. Skrupa, 372 U. S. 726, 731-732 (1963); Usery v. Turner Elkhorn Mining Co., 428 U. S. 1, 15 (1976). Appellees, however, urge a more elevated standard of review on the ground that the interests jeopardized by the Price-Anderson Act “are far more important than those in the economic due process and business-oriented cases” where the traditional rationality standard has been invoked. Brief for Appellees 36. An intermediate standard like that applied in cases such as Craig v. Boren, 429 U. S. 190 (1976) (equal protection challenge to statute requiring that males be older than females in order to purchase beer) or United States Trust Co. of New York v. New Jersey, 431 U. S. 1 (1977) (Contract Clause challenge to repeal of statutory covenant providing security for bondholders) is thus recommended for our use here.

As we read the Act and its legislative history, it is clear that Congress’ purpose was to remove the economic impediments in order to stimulate the private development of electric energy by nuclear power while simultaneously providing the public compensation in the event of a catastrophic nuclear incident. See, e. g., S. Rep. No. 296, 85th Cong., 1st Sess., 15 (1957). The liability-limitation provision thus emerges as a classic example of an economic regulation — a legislative effort to structure and accommodate “the burdens and benefits of economic life.” Usery v. Turner Elkhorn Mining Co., supra, at 15. “It is by now well established that [such] legislative Acts . . . come to the Court with a presumption of constitutionality, and that the burden is on one complaining of a due process violation to establish that the legislature has acted in an arbitrary and irrational way.” Ibid. That the accommodation struck may have profound and far-reaching consequences, contrary to appellees’ suggestion, provides all the *84more reason for this Court to defer to the congressional judgment unless it is demonstrably arbitrary or irrational.27

B

When examined in light of this standard of review, the Price-Anderson Act, in our view, passes constitutional muster. The record before us fully supports the need for the imposition of a statutory limit on liability to encourage private industry participation and hence bears a rational relationship to Congress’ concern for stimulating the involvement of private enterprise in the production of electric energy through the use of atomic power; nor do we understand appellees or the District Court to be of a different view. Rather their challenge is to the alleged arbitrariness of the particular figure of $560 million, which is the statutory ceiling on liability. The District Court aptly summarized its position:

“The amount of recovery is not rationally related to the potential losses. Abundant evidence in the record shows that although major catastrophe in any particular place is not certain and may not be extremely likely, nevertheless, in the territory where these plants are located, damage to life and property for this and future generations could well be many, many times the limit which the law places on liability.” 431 F. Supp., at 222.

Assuming, arguendo, that the $560 million fund would not insure full recovery in all conceivable circumstances28 — and *85the hard truth is that no one can ever know — it does not by any means follow that the liability limitation is therefore irrational and violative of due process. The legislative history clearly indicates that the $560 million figure was not arrived at on the supposition that it alone would necessarily be sufficient to guarantee full compensation in the event of a nuclear incident. Instead, it was conceived of as a “starting point” or a working hypothesis.29 The reasonableness of the statute's assumed ceiling on liability was predicated on two corollary considerations — expert appraisals of the exceedingly small risk of a nuclear incident involving claims in excess of $560 million, and the recognition that in the event of such an incident, Congress would likely enact extraordinary relief provisions to provide additional relief, in accord with prior practice.

“[T]his limitation does not, as a practical matter, detract from the public protection afforded by this legislation. In the first place, the likelihood of an accident occurring *86which would result in- claims exceeding the sum of the financial protection required and the governmental indemnity is exceedingly remote, albeit theoretically possible. Perhaps more important, in the event of a national disaster of this magnitude, it is obvious that Congress would have to review the problem and take appropriate action. The history of other natural or man-made disasters, such as the Texas City incident, bears this out. The limitation of liability serves primarily as a device for facilitating further congressional review of such a situation, rather than as an ultimate bar to further relief of the public.” H. R. Rep. No. 883, 89th Cong., 1st Sess., 6-7 (1965).

See also S. Rep. No. 296, supra, at 21; H. R. Rep. No. 9-M348, pp. 12, 15 (1975).

Given our conclusion that, in general, limiting liability is an acceptable method for Congress to utilize in encouraging the private development of electric energy by atomic power, candor requires acknowledgment that whatever ceiling figure is selected will, of necessity, be arbitrary in the sense that any choice of a figure based on imponderables like those at issue here can always be so characterized. This is not, however, the kind of arbitrariness which flaws otherwise constitutional action. When appraised in terms of both the extremely remote possibility of an accident where liability would exceed the limitation 30 and Congress’ now statutory commitment to “take whatever action is deemed necessary and appropriate to protect the public from the consequences of” any such disaster, 42 U. S. C. § 2210 (e) (1970 ed., Supp. V),31 we hold the *87congressional decision to fix a $560 million ceiling, at this stage in the private development and production of electric energy by nuclear power, to be within permissible limits and not violative of due process.

This District Court’s further conclusion that the Price-Anderson Act “tends to encourage irresponsibility ... on the part of builders and owners” of the nuclear power plants, 431 F. Supp., at 222, simply cannot withstand careful scrutiny. We recently outlined the multitude of detailed steps involved in the review of any application for a license to construct or to operate a nuclear power plant, Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U. S. 519, 526-527, and n. 5 (1978); nothing in the liability-limitation provision undermines or alters in any respect the rigor and integrity of that process. Moreover, in the event of a nuclear accident the utility itself would suffer perhaps the largest damages. While obviously not to be compared with the loss of human life and injury to health, the. risk of financial loss and possible bankruptcy to the utility is in itself no small incentive to avoid the kind of irresponsible and cavalier conduct implicitly attributed to licensees by the District Court.

The remaining due process objection to the liability-limitation provision is that it fails to provide those injured by a *88nuclear accident with a satisfactory quid pro quo for the common-law rights of recovery which the Act abrogates. Initially, it is not at all clear that the Due Process Clause in fact requires that a legislatively enacted compensation scheme either duplicate the recovery at common law or provide a reasonable substitute remedy.32 However, we need not resolve this question here since the Price-Anderson Act does, in our view, provide a reasonably just substitute for the common-law or state tort law remedies it replaces. Cf. New York Central R. Co. v. White, 243 U. S. 188 (1917) ’; Crowell v. Benson, 285 U. S, 22 (1932).33

*89The legislative history of the hability-limitation provisions and the accompanying compensation mechanism reflects Congress’ determination that reliance on state tort law remedies and state-court procedures was an unsatisfactory approach to assuring public compensation for nuclear accidents, while at the same time providing the necessary incentives for private development of nuelear-produced energy. The remarks of Chairman Anders of the NRC before the Joint Committee on Atomic Energy during the 1975 hearings on the need for renewal of the Price-Anderson Act are illustrative of this concern and of the expectation that the Act would provide a more efficient and certain vehicle for assuring compensation in the unlikely event of a nuclear incident:

“The primary defect of this alternative [nonrenewal of the Act], however, is its failure to afford the public either a secure source of funds or a firm basis for legal liability with respect to new plants. While in theory no legal limit would be placed on liability, as a practical matter the public would be less assured of obtaining compensation than under Price-Anderson. Establishing liability would depend in each case on state tort law and procedures, and these might or might not provide for no-fault liability, let alone the multiple other protections now embodied in Price-Anderson. The present assurance of prompt and equitable compensation under a pre-struc-tured and nationally applicable protective system would give way to uncertainties, variations and potentially lengthy delays in recovery. It should be emphasized, moreover, that it is collecting a judgment, not filing a *90lawsuit, that counts. Even if defenses are waived under state law, a defendant with theoretically “unlimited” liability may be unable to pay a judgment once obtained. When the defendant’s assets are exhausted by earlier judgments, subsequent claimants would be left with un-collectable awards. The prospect of inequitable distribution would produce a race to the courthouse door in contrast to the present system of assured orderly and equitable compensation.” Hearings on H. R. 8631 before Joint Committee on Atomic Energy, 94th Cong., 1st Sess., 69 (1975).

Appellees, like the District Court, differ with this appraisal on several grounds. They argue, inter alia,* that recovery under the Act would not be greater than without it, that the waiver of defenses required by the Act, 42 U. S. C. § 2210 (n) (1970 ed., Supp. V), is an idle gesture since those involved in the development of nuclear energy would likely be held strictly liable under common-law principles;34 that the claim-administration procedure under the Act delays rather than expedites individual recovery; and finally that recovery of even limited compensation is uncertain since the liability ceiling does not vary with the number of persons injured or amount of property damaged. The extension of short state statutes of limitations and the provision of omnibus35 coverage do not save the Act, in their view, since such provisions could equally well be included in a fairer plan which would assure greater compensation.

We disagree. We view the congressional assurance of a $560 million fund for recovery, accompanied by an express statutory commitment, to “take whatever action is deemed necessary *91and appropriate to protect the public from the consequences of” a nuclear accident, 42 U. S. C. § 2210 (e) (1970 ed., Supp. V), to be a fair and reasonable substitute for the uncertain recovery of damages of this magnitude from a utility or component manufacturer, whose resources might well be exhausted at an early stage. The record in this case raises serious questions about the ability of a utility or component manufacturer to satisfy a judgment approaching $560 million — the amount guaranteed under the Price-Anderson Act.36 Nor are we persuaded that the mandatory waiver of defenses required by the Act is of no benefit to potential claimants. Since there has never been, to our knowledge, a case arising out of a nuclear incident like those covered by the Price-Anderson Act, any discussion of the standard of liability that state courts will apply is necessarily speculative. At the minimum, the statutorily mandated waiver of defenses establishes at the threshold the right of injured parties to compensation without proof of fault and eliminates the burden of delay and uncertainty which would follow from the need to litigate the question of liability after an accident. Further, even if strict liability were routinely applied, the common-law doctrine is subject to exceptions for acts of God or of third parties37 — two of the very factors which appellees emphasized in the District Court in *92the course of arguing that the risks of a nuclear accident are greater than generally admitted. All of these considerations belie the suggestion that the Act leaves the potential victims of a nuclear disaster in a more disadvantageous position than they would be in if left to their common-law remedies — not known in modern times for either their speed or economy.

Appellees’ remaining objections can be briefly treated. The claim-administration procedures under the Act provide that in the event of an accident with potential liability exceeding the $560 million ceiling, no more than 15% of the limit can be distributed pending court approval of a plan of distribution taking into account the need to assure compensation for “possible latent injury claims which may not be discovered until a later time.” 42 U. S. C. §2210 (o)(3) (1970 ed., Supp. V). Although some delay might follow from compliance with this statutory procedure, we doubt that it would approach that resulting from routine litigation of the large number of claims caused by a catastrophic accident.38 Moreover, the statutory scheme insures the equitable distribution of benefits to all who suffer injury' — both immediate and latent; under the common-law route, the proverbial race to the courthouse would instead determine who had “first crack” at the diminishing resources of the tortfeasor, and fairness could well be sacrificed in the process. The remaining contention that recovery is uncertain because of the aggregate rather than individualized nature of the liability ceiling is but a thinly disguised version of the contention that the $560 million figure is inadequate, which we have already rejected.

In the course of adjudicating a similar challenge to the *93Workmen’s Compensation Act in New York Central R. Co. v. White, 243 U. S., at 201, the Court observed that the Due Process Clause of the Fourteenth Amendment was not violated simply because an injured party would not be able to recover as much under the Act as before its enactment. “ [H] e is entitled to moderate compensation in all cases of injury, and has a certain and speedy remedy without the difficulty and expense of establishing negligence or proving the amount of the damages.” The logic of New York Central would seem to apply with renewed force in the context of this challenge to the Price-Anderson Act. The Price-Anderson Act not only provides a reasonable, prompt, and equitable mechanism for compensating victims of a catastrophic nuclear incident, it also guarantees a level of net compensation generally exceeding that recoverable in private litigation. Moreover, the Act contains an explicit congressional commitment to take further action to aid victims of a nuclear accident in the event that the $560 million ceiling on liability is exceeded. This panoply of remedies and guarantees is at the least a reasonably just substitute for the common-law rights replaced by the Price-Anderson Act. Nothing more is required by the Due Process Clause.

Although the District Court also found the Price-Anderson Act to contravene the “equal protection provision that is included within the Due Process Clause of the Fifth Amendment,” 431 F. Supp., at 224-225, appellees have not relied on this ground since the equal protection arguments largely track and duplicate those made in support of the due process claim. In any event, we conclude that there is no equal protection violation. The general rationality of the Price-Anderson Act liability limitations — particularly with reference to the important congressional purpose of encouraging private participation in the exploitation of nuclear energy — is ample justification for the difference in treatment between those injured in nuclear accidents and those whose injuries are derived from other *94causes. Speculation regarding other arrangements that might be used to spread the risk of liability in ways different from the Price-Anderson Act is, of course, not pertinent to the equal protection analysis. See Mourning v. Family Publications Service, Inc., 411 U. S. 356, 378 (1973).39

Accordingly, the decision of the District Court is reversed, and the cases are remanded for proceedings consistent with this opinion.

Reversed and remanded.

Mr. Justice Stewart,

concurring in the result.

With some difficulty I can accept the proposition that federal subject-matter jurisdiction under 28 U. S. C. § 1331 (1976 ed.) exists here, at least with respect to the suit against the Nuclear Regulatory Commission, the agency responsible for the administration of the Price-Anderson Act. The claim under federal law is to be found in the allegation that the Act, if enforced, will deprive the appellees of certain property rights, in violation of the Due Process Clause of the Pifth Amendment. One of those property rights, and perhaps the sole cognizable one, is a state-created right to recover full compensation for tort injuries. The Act impinges on that right by limiting recovery in major accidents.

*95But there never has been such an accident, and it is sheer speculation that one will ever occur. For this reason I think there is no present justiciable controversy, and that the appel-lees were without standing to initiate this litigation.

On the issue of standing, the Court relies on the “present” injuries of increased water temperatures and low-level radiation emissions. Even assuming that but for the Act the plant would not exist and therefore neither would its effects on the environment, I cannot believe that it follows that the appel-lees have standing to attack the constitutionality of the Act. Apart from a “but for” connection in the loosest sense of that concept, there is no relationship at all between the injury alleged for standing purposes and the injury alleged for federal subject-matter jurisdiction.

Surely a plaintiff does not have standing simply because his challenge, if successful, will remove the injury relied on for standing purposes only because it will put the defendant out of existence. Surely there must be some direct relationship between the plaintiff’s federal claim and the injury relied on for standing. Cf. Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 261; United States v. SCRAP, 412 U. S. 669, 687-690; Linda R. S. v. Richard D., 410 U. S. 614, 617-618. An interest in the local water temperature does not, in short, give these appellees standing to bring a suit under 28 U. S. C. § 1331 (1976 ed.) to challenge the constitutionality of a law limiting liability in an unrelated and as-yet-to-occur major nuclear accident.

For these reasons, I would remand these cases to the District Court with instructions to dismiss the complaint.

Mr. Justice Rehnquist,

with whom Mr. Justice Stevens joins, concurring in the judgment.

I can understand the Court’s willingness to reach the merits of this case and thereby remove the doubt which has been cast over this important federal statute. In so doing, however, it ignores established limitations on district court jurisdiction *96as carefully defined in our statutes and cases. Because I believe the preservation of these limitations is in the long run more important to this Court’s jurisprudence than the resolution of any particular case or controversy, however important, I too would reverse the judgment of the District Court, but would do so with instructions to dismiss the complaint for want of jurisdiction. Cf. Montana-Dakota Utilities Co. v. Northwestern Public Service Co., 341 U. S. 246, 249-250 (1951).

Giving the conclusory allegations of appellees’ complaint the most liberal possible reading, they purport to establish only two grounds for the declaratory relief requested.. First, they contend that the Price-Anderson Act deprives them of their property without due process of law in that it irrationally limits the tort recovery otherwise available in the North Carolina courts.1 Second, they contend that the Act works an unconstitutional taking of their property for public use without just compensation. They purport to base District Court jurisdiction upon 28 U. S. C. § 1337 (1976 ed.) which covers “any civil action or proceeding arising under any Act of Congress regulating commerce or protecting trade and commerce against restraints and monopolies.”

I

It is apparent that appellees’ first asserted basis for relief does not state a claim “arising under” the Price-Anderson Act. Their complaint alleges that the operation of the two power plants will cause immediate injury to property within their vicinity. App. 32, ¶ 21. The District Court explicitly found that these injuries “give rise to an immediate right of action for redress. Under the law of North Carolina a right of action arises as soon as a wrongful act has created 'any injury, how*97ever slight/ to the plaintiff.” 431 F. Supp. 203, 221 (WDNC 1977) (citations omitted). This right of action provided by state, not federal, law is the property of which the appellees contend the Act deprives them without due process. Thus, the constitutionality of the Act becomes relevant only if the appellant Duke Power Co; were to invoke the Act as a defense to appellees’ suit for recovery under their North Carolina right of action.

It has long been established that the mere anticipation of a possible federal defense to a state cause of action is not sufficient to invoke the federal-question jurisdiction of the district courts. In Louisville & Nashville R. Co. v. Mottley, 211 U. S. 149 (1908), the plaintiffs sought to compel the specific performance of a contract by which the railroad had. granted them free passes for life. Although their contract was not predicated upon federal law, the plaintiffs contended that federal-question jurisdiction was established by the presence of an Act of Congress forbidding railroads to issue free passes. This Court held that the District Court did not have jurisdiction to consider whether the Act was inapplicable or unconstitutional:

“It is the settled interpretation of these words [‘arising under’], as used in this statute, conferring jurisdiction, that a suit arises under the Constitution and laws of the United States only when the plaintiff’s statement of his own cause of action shows that it is based upon those laws or that Constitution. It is not enough that the plaintiff alleges some anticipated defense to his cause of action and asserts that the defense is invalidated by some provision of the Constitution of the United States. Although such allegations show that very likely, in the course of the litigation, a question under the Constitution would arise, they do not show that the suit, that is, the plaintiff’s original cause of action, arises under the Constitution.” Id., at 152.

*98Just as the underlying claim in Mottley arose under Kentucky contract law, the underlying claim in this case arises under North Carolina tort law. This Court has construed the “arising under” language of 28 U. S. C. § 1337 (1976 ed.) just as it has the similar language of 28 U. S. C. § 1331 (1976 ed.). Peyton v. Railway Express Agency, Inc., 316 U. S. 350, 353 (1942).

Nor does the fact that appellees seek only declaratory relief under the Declaratory Judgment Act, 28 U. S. C. § 2201 (1976 ed.), support a different result. This Court has held that the well-pleaded complaint rule applied in Mottley is fully applicable in cases seeking only declaratory relief, because the Declaratory Judgment Act merely expands the remedies available in the district courts without expanding their jurisdiction. “It would turn into the federal courts a vast current of litigation indubitably arising under State law, in the sense that the right to be vindicated was State-created, if a suit for a declaration of rights could be brought into the federal courts merely because an anticipated defense derived from federal law.” Skelly Oil Co. v. Phillips Petroleum Co., 339 U. S. 667, 673 (1950). See also Phillips Petroleum Co. v. Texaco Inc., 415 U. S. 125 (1974); C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 3566, pp. 437-438 (1975).2

*99Appellees do not contend that the Price-Anderson Act itself grants to them personal rights which may be vindicated in a federal proceeding. Since the only property rights they assert arise under North Carolina law, the District Court had no jurisdiction to consider whether the setting up of an Act of Congress as a defense against those rights would deny them due process of law under the Fifth Amendment.

Indeed, the Court does not even contend that there is an independent statutory source of jurisdiction over Duke. Ante, at 72 n. 16. It suggests instead that the complaint states a claim against the Nuclear Regulatory Commission, not as a joint tortfeasor under North Carolina law, but as the administrator of an unconstitutional federal statute. The Court’s theory is that the complaint alleges the existence of an implied right of action under the Fifth Amendment to obtain relief against arbitrary federal statutes. It can hardly be said that this theory of the case emerges with crystal clarity from either the complaint or the brief of the appellees.

More importantly, there is no allegation in this complaint that the Nuclear Regulatory Commission has taken or will take any unconstitutional action at all. The complaint alleges only that the Commission granted construction permits to *100Duke, and that it will enter into an agreement “to indemnify Duke for any nuclear incident exceeding the amount of $125,000,000 subject to a maximum liability of $560,000,000.” App. 31, ¶ 13. Neither of these actions is alleged to be unconstitutional. The gist of the complaint is the asserted unconstitutionality of 42 U. S. C. § 2210 (e) (1970 ed., Supp. V), which limits Duke’s liability. But this limitation of liability is separate and apart from the indemnity agreement which the Commission is authorized to execute under 42 U. S. C. § 2210 (d) (1970 ed., Supp. Y). The Commission has nothing whatever to do with the administration of the limitation of liability; whatever administration of that statute there is to be is left in the hands of the District Court. 42 U. S. C. §2210 (o) (1970 ed. and Supp. V). The District Court, of course, is not a party to this suit.3

It simply cannot be said that these allegations make out an actual controversy against the Commission. While the Commission may be quite interested in the constitutionality of the statute, that is hardly sufficient to establish a justiciable controversy. Muskrat v. United States, 219 U. S. 346, 361-362 (1911). While appellees may have been damaged by Duke’s decision to construct these plants, there is no “challenged action of the defendant” Commission to which their damage “fairly can be traced.” Simon v. Eastern Ky. Wel *101 fare Rights Org., 426 U. S. 26, 41 (1976). If Duke decided to proceed with construction despite a declaration of the statute’s unconstitutionality, there would be nothing that the Commission could do to aid appellees. Where the prospect of effective relief against a defendant depends on the actions of a third party, no justiciable controversy exists against that defendant. Warth v. Seldin, 422 U. S. 490, 505 (1975). In short, appellees’ only conceivable controversy is with Duke, over whom the District Court had no jurisdiction.

II

As appellees themselves describe the second aspect of their complaint, “the central issue is whether in the circumstances of this case, the complete destruction of appellees’ property by a nuclear accident, occurring at one of Duke’s plants, would be a ‘taking’ by the United States, as that term is defined in the Fifth Amendment.” Brief for Appellees 62. This statement makes clear that appellees’ claim arises not under the Price-Anderson Act but under the Fifth Amendment itself. Jurisdiction under § 1337 extends only to actions vindicating rights created by an Act of Congress. Compare Switchmen v. National Mediation Board, 320 U. S. 297, 300 (1943), with General Committee v. Missouri-Kansas-Texas R. Co., 320 U. S. 323, 337 (1943). Since it cannot be maintained that the Price-Anderson Act created appellees’ asserted right to be free from takings for public use without just compensation, it follows that District Court jurisdiction may not be predicated upon § 1337.

The District Court does have jurisdiction to consider claims of taking under the Tucker Act, 28 U. S. C. § 1346 (a) (2) (1976 ed.), where the amount in controversy does not exceed $10,000.4 “But the Act has long been construed as authoriz*102ing only actions for money judgments and not suits for equitable relief against the United States.” Richardson v. Morris, 409 U. S. 464, 465 (1973). It is incontrovertibly established that neither the Court of Claims nor the district courts have jurisdiction under the Tucker Act to issue the sort of declaratory relief granted here. Compare ibid., with United States v. King, 395 U. S. 1 (1969). Thus, the record does not establish any jurisdictional basis upon which the District Court could grant declaratory relief on appellees’ taking claim.

There being no basis for District Court jurisdiction over either of appellees’ claims, its judgment should be reversed and the cause remanded with instructions to dismiss the complaint for want of jurisdiction.

Mr. Justice Stevens,

concurring in the judgment.

The string of contingencies that supposedly holds this litigation together is too delicate for me. We are told that but for the Price-Anderson Act there would be no financing of nuclear power plants, no development of those plants by private parties, and hence no present injury to persons such as appel-lees; we are then asked to remedy an alleged due process viola*103tion that may possibly occur at some uncertain time in the future, and may possibly injure the appellees in a way that has no significant connection with any present injury. It is remarkable that such a series of speculations is considered sufficient either to make this litigation ripe for decision or to establish appellees’ standing;* it is even more remarkable that this occurs in a case in which, as Mr. Justice Rehnquist demonstrates, there is no federal jurisdiction in the first place.

The Court’s opinion will serve the national interest in removing doubts concerning the constitutionality of the Price-Anderson Act. I cannot, therefore, criticize the statesmanship of the Court’s decision to provide the country with an advisory opinion on an important subject. Nevertheless, my view of the proper function of this Court, or of any other federal court, in the structure of our Government is more limited. We are not statesmen; we are judges. When it is necessary to resolve a constitutional issue in the adjudication of an actual case or controversy, it is our duty to do so. But whenever we are persuaded by reasons of expediency to engage in the business of giving legal advice, we chip away a part of the foundation of our independence and our strength.

I join Mr. Justice Rehnquist’s opinion concurring in the judgment.