Opinion for the Court by HARRY T. EDWARDS, Circuit Judge.
HARRY T. EDWARDS, Circuit Judge:
In this case, the petitioners challenge the legality of the Hazardous Ranking System (“HRS”), adopted by the Environmental Protection Agency (“EPA” or the “agency”) pursuant to section 105 of the Comprehensive Environmental Response, Compensation and Liability Act of 1980 (“CERCLA”).1 The HRS is the designated model used by the EPA to determine sites to be listed on a National Priority List (“NPL”) under CERCLA. The NPL lists 400 sites that have been contaminated by harmful substances and that may warrant corrective action under CERCLA. Petitioners here claim that they should not have been included on the NPL because the ranking methodology established by the HRS is unlawful.2
The HRS was promulgated on July 16, 1982, in a notice-and-comment rulemaking proceeding separate from that which produced the final NPL on September 8, 1983. CERCLA’s statutory review provision, section 113(a),3 stipulates that petitions for judicial review of regulations pro[183]*183mulgated under CERCLA must be filed with this court within ninety days of the regulations’ promulgation. The petitioners failed to seek review of the HRS during the mandated statutory period. Despite their untimely request for review, the petitioners urge that their failure tó file within the statutory period should be excused because they assumed that their challenge was not ripe for review until the NPL was promulgated. We reject this argument as patently untenable under the applicable terms of CERCLA.
We emphasize first that petitioners who delay filing requests for review on their own assessment of when an issue is ripe for review do so at the risk of finding their claims time-barred. Normally, the appropriate time for a judicial determination of the ripeness of an issue is within the prescribed statutory period for review. In general, we will refuse to hypothesize whether, in retrospect, a claim would have been deemed ripe for review had it been brought during the statutory period, in order to save an untimely claim. Exceptions occasionally may be justified in the light of changed circumstances giving rise to a new cause of action beyond the statutory period for review; compelling case precedent that makes it clear beyond doubt that the claim was not ripe during the statutory period; or clear evidence that a failure to consider a petitioner’s claims would work a manifest injustice. The petitioners’ challenge to the HRS fall within none of these exceptions. However, because we articulate here for the first time the circumstances under which the court will engage in “retrospective ripeness analysis” after the statutory review period has expired, we proceed to evaluate the petitioners’ claim on the merits.
We conclude that the challenge to the HRS was ripe during the statutory review period. Therefore, we find petitioners’ claim to be barred as untimely. In the alternative, we hold on the merits that, given the narrow purpose of the HRS and the NPL — to provide an expeditious and relatively inexpensive initial determination of which sites may warrant further action under CERCLA — and in light of the agency’s manifest awareness of the HRS’s technical limitations, the model is reasonable and consistent with congressional intent.
I. Background
Through CERCLA, Congress sought to establish a system for rectifying some of the serious public health and environmental problems that have been caused by improper disposal of hazardous wastes, pollutants and contaminants. Essentially, CERCLA authorizes the EPA to respond to actual or threatened releases of these harmful substances,4 both through removal actions, which entail the actual cleanup of a release,5 and remedial actions, which provide for remedies to prevent or minimize the release of hazardous substances.6 The statute also establishes a fund (the “Superfund”) to pay for cleanup action.7
In section 1058 of CERCLA, Congress instructs the EPA9 to revise the National Contingency Plan (“NCP”), which had originally been developed under the Clean Water Act10 to give guidance to agencies in removing oil or hazardous substances from United States waters, to “reflect and effectuate the responsibilities and powers” created by CERCLA. This revision is to in-[184]*184elude “criteria for determining priorities among releases or threatened releases throughout the United States____”11 The EPA is also directed to include in the NCP an initial list of at least 400 sites selected on the basis of the above criteria — the NPL. The statute requires that the NPL be revised “no less often than annually.” 12
In response to this statutory mandate, the EPA adhered to the requisites of notice-and-comment rulemaking and published a proposed revision of the NCP in the Federal Register on March 12, 1982.13 This proposal explicitly designated the HRS as the method favored by the agency for ranking hazardous substance releases, included an address where the HRS could be obtained upon request, and solicited comment on the HRS.14 Numerous comments on the proposed HRS were received and considered by the EPA during the rulemaking process. The final versions of the revised NCP and the HRS were then promulgated on July 16, 1982, and published in the Federal Register.15
As the EPA explained in the preamble to the final NCP, “the HRS is designed to estimate the potential hazard presented by releases or threatened releases of hazardous substances, pollutants and contaminants.” 16 The agency applies the HRS to data from an observed or potential release to obtain a “score” or estimate of the risk from the release. The EPA then relies on HRS scores to determine which releases should be listed on the NPL.17
The HRS score is calculated by analyzing three potential “pathways” of human or environmental exposure: (1) ground water, (2) surface water, and (3) air. Within each of these pathways, the HRS takes account of three categories of factors “that are designed to encompass most aspects of the likelihood of exposure to a hazardous substance through a release and the magnitude or degree of harm from such exposure.” 18 These categories reflect: “(1) [t]he existence or likelihood of a release, (2) the characteristics of the hazardous substances that have been or may be released, and (3) the population or sensitive environment that is threatened.” 19 Each of these three categories incorporates a number of separate factors that receive numerical values according to a set scale. For instance, the factors that are analyzed and assigned numerical values under category “(2)” include the toxicity, persistence and quantity of the substance. After numerical values are computed for each factor, the HRS employs mathematical formulas that “reflect the relative importance and interrelationships of the various factors”20 to arrive at a final score. The resultant HRS score represents an estimate of “the probability and magnitude of harm to the human population or sensitive environment from exposure to hazardous substances as a result of [185]*185contamination of ground water, surface water, or air.”21
The EPA applied the HRS to hundreds of sites, yielding a list of scores that was used to create the NPL.22 The proposed NPL was published on December 30,1982;23 the final version was promulgated as a rule nine months later on September 8, 1983.24
The major purpose of the NPL and the HRS, as the EPA explained in the preamble to the final version of the NPL, is narrowly focused. It is to identify, quickly and inexpensively, sites that may warrant further action under CERCLA.25 Listing does not represent a determination that action is necessary, or that the EPA will take action.26
With this background information in mind, we turn now to the petitioners’ challenge to the legality of the HRS.
II. The Timeliness of the Petitioners’ Challenge to the HRS
A. Introduction
~We first consider whether the petitioners’ challenge is barred because it was not filed within the statutory review period.
Section 113(a) of CERCLA provides that [r]eview of any regulation promulgated under [CERCLA] may be had upon application by any interested person only in the Circuit Court of Appeals of the United States for the District of Columbia. Any such application shall be made within ninety days from the date of promulgation of such regulations. Any matter with respect to which review could have been obtained under this subsection shall not be subject to judicial review in any civil or criminal proceeding for enforcement or to obtain damages or recovery of response costs.27
This court has repeatedly recognized that statutory time limits on petitions for review of agency action are jurisdictional in nature.28 These limitations serve “the important purpose of imparting finality into the administrative process, thereby conserving administrative resources.”29 Furthermore, timeliness requirements reflect “a deliberate congressional choice to impose statutory finality on agency orders, a choice we may not second-guess.”30 We have entertained untimely claims only in a limited number of exceptional circumstances where the petitioner lacked a meaningful opportunity to challenge the agency action during the review period due to, for example, inadequate notice that the petitioner would be affected by the action,31 confusion in the law as to the proper forum for [186]*186review,32 and lack of ripeness during the review period.33 Proffered excuses for late filing are carefully scrutinized for, as we have noted, “the court must be strict in determining what constitutes a ‘legitimate' excuse [for failure to file a timely petition]; otherwise, the policy of finality underlying the [time] limit will not be achieved.”34
The petitioners here admit that they did not seek judicial review of the HRS during the ninety day period following its promulgation. The EPA maintains that the petitioners could have challenged the HRS during the statutory review period and that consequently present review is barred. The petitioners contend that their complaint about the HRS did not become ripe for review until the NPL was promulgated and that, because they filed a petition to review the NPL within ninety days of its promulgation, their challenge to the HRS is also timely. In support of their position, the petitioners urge that our precedents, in particular Diamond Shamrock Corporation v. Costle,35 establish that agency rule-making that creates a standard, such as the effluent limitation regulations in Diamond Shamrock and the HRS here, is never ripe for review — and thus the statutory petition period is tolled — until the agency applies the rule to the petitioner. Under this theory, the petitioners’ objections to the HRS ripened only when, in order to develop the NPL, the EPA applied the HRS to sites with which the petitioners are associated. Counsel for the petitioners also seemed to suggest at oral argument that Diamond Shamrock can be read to waive the explicit timely filing requirement contained in section 113(a) of CERCLA.
B. The Relationship Between Ripeness and Timeliness
As an initial matter, we think that the petitioners confound the obligations of the court with those of the petitioner. It is the duty of the court to make the prudential judgment whether a challenge to agency action is ripe; it is the responsibility of petitioners to file for review within the period set by Congress. While it is true that in extraordinary circumstances we have “forgiven” a petitioner’s failure to file a timely petition, we have never suggested that petitioners may safely substitute their own notions of timely review for those of Congress. To the contrary, we have previously admonished petitioners of the wisdom of filing protective petitions for review during the statutory period,36 and we do so again now. Diamond Shamrock,37 so heavily relied upon by petitioners, is inapposite. In that case we simply upheld the District Court’s decision that a complaint seeking review of some effluent limitation regulations issued by the EPA was not yet ripe. The court did not consider the relationship between ripeness and timeliness at all.
Considering that relationship here, we think it certainly is not the one implied by the petitioners — a rather casual means of circumventing timely filing requirements. Ripeness “is concerned primarily with the institutional relationships between courts and agencies, and the competence of the courts to resolve disputes without further administrative refinement of the issues.” 38 It is largely a prudential doctrine, whose “basic rationale,” as the Supreme Court explained in the landmark decision, Abbott Laboratories v. Gardner, [187]*187“is to prevent the courts, through avoidance of premature adjudication, from entangling themselves in abstract disagreements over administrative policies, and also to protect the agencies from judicial interference until an administrative decision has been formalized and its effects felt in a concrete way by the challenging parties.”39 Timeliness requirements, on the other hand, are designed to protect a different set of interests. They are intended to prevent courts from “entangling themselves” in disputes which Congress has determined have been raised too late and to protect agencies from endless judicial interference with formalized administrative policy.40
Because the principal function of the ripeness doctrine is to aid a court in ascertaining whether it should stay its hand until agency policy has crystallized, most of the case law explicating the doctrine is forward-looking, that is, looking into the future to determine the effects of deferring review.41 Only rarely does case law depict the ripeness doctrine in the service of a tardy petitioner — looking backward to divine whether the court would have considered the request for review ripe had it been brought in a timely fashion. Indeed, we know of only one case from this circuit, Getter v. FCC,42 where we explicitly evaluated the retrospective ripeness of an úntimely challenge.43 There the petitioner solicited review of an FCC regulation nearly five years after the statutory review period had expired. We dismissed all of Geller’s claims for lack of timeliness, except one which had clearly just become ripe with the passage of new legislation.44
We think Getter may suggest one of two kinds of cases where the court should perform a retrospective ripeness analysis for a petitioner who has failed to file a timely request for review. The Getter type can be characterized as a case in which events occur or information becomes available after the statutory review period expires that essentially create a challenge that did not previously exist.45 The other type involves [188]*188claims that, under our precedents, are without any doubt not ripe for review during the statutory period. For example, Diamond Shamrock held that certain effluent limitation regulations promulgated under the Federal Water Pollution Control Act (“FWPCA”)46 were not ripe for review until applied in discharge permit proceedings.47 Obviously, other would-be petitioners in the same position as those in Diamond Shamrock —applicants for discharge permits under the same statute— can rely on the holding in that case.
As a general proposition, however, if there is any doubt about the ripeness of a claim, petitioners must bring their challenge in a timely fashion or risk being barred. Courts simply are not well-suited to answering hypothetical questions which involve guessing what the court might have done in the past. Furthermore, if we were routinely to conduct retrospective ripeness analyses where a late petitioner offers no compelling justification for not having filed his claim in a timely manner, we would wreak havoc with the congressional intention that repose be brought to final agency action. Consequently, except where events occur or information becomes available after the statutory review period expires that essentially create a challenge that did not previously exist, or where a petitioner’s claim is, under our precedents, indisputably not ripe until the agency takes further action, we will be very reluctant, in order to save a late petitioner from the strictures of a timeliness requirement, to engage in a retrospective determination of whether we would have held the claim ripe had it been brought on time.
We do not believe that the instant case falls into either the Geller or clear precedent exceptions. It is not a Geller case because no events occurred after the statutory period that gave rise to an essentially new claim. Where, as in this case, the petitioner has notice during the review period that the regulation pertains to him,48 the mere fact that the rule is applied to the petitioner after the statutory period expires will normally not be sufficient to bring a case into the Geller category. The delayed application of the rule may be relevant, however, to the clear precedent category. For example, as we noted above, other petitioners in Diamond Shamrock’s position could rely on our holding that the regulations at issue in that case were not ripe until applied.
The petitioners argue that Diamond Shamrock also governs the ripeness issue presented by this case. We disagree. Diamond Shamrock involved “pre-enforcement” review by the District Court of regulations that would have been directly reviewable in this court under the review provision of the FWPCA upon the application of the regulations to the petitioner in a permit proceeding.49 There, the court explicitly found that both the agency and the court would benefit from deferring judicial consideration until review under the organic statute was available.50 By contrast, review here is sought under CERCLA’s review provision, which expresses a congressional intent that review of regulations immediately follow promulgation and that it not take place at the enforcement stage. Furthermore, as we explain more fully in the succeeding section of this opinion, we find that both the EPA and the court would have benefitted from conducting the requested review of the HRS within the statutory period. Thus, Diamond Shamrock [189]*189is in no way dispositive of the instant case.51
Generally, if we determine that a petitioner’s request for a retrospective ripeness evaluation does not fit into either the Geller or the clear precedent categories we will refuse to examine the petitioner's untimely claim to ascertain whether we would have found it ripe for adjudication had it been brought within the statutory period, unless our refusal to do so would cause a serious injustice to the petitioner. However, because we clearly articulate this policy for the first time in this case, we pause to demonstrate that the petitioners’ challenge to the HRS was ripe during the review period.
C. Ripeness Analysis
Ripeness law overlaps at its borders with Article III requirements of case or controversy.52 However, the primary focus of the ripeness doctrine as it concerns judicial review of agency action has been a prudential attempt to time review in a way that balances the petitioner’s interest in prompt consideration of allegedly unlawful agency action against the agency’s interest in crystallizing its policy before that policy is subjected to judicial review and the court’s interests in avoiding unnecessary adjudication and in deciding issues in a concrete setting.53
In Abbott Laboratories, the Supreme Court announced a twofold test for balancing these interests in order to determine the ripeness' of “pre-enforcement” agency actions for judicial review. That test requires an evaluation of “the fitness of the issues for judicial decision and the hardship to the parties of withholding court consideration.” 54
Under the “fitness of the issues” prong, this court takes account of the institutional capacities of, and the relationship between, courts and agencies. We look, for example, to see if the issue raises a purely legal question. If it does, we assume its threshold suitability for judicial determination.55 We also consider whether the agency or the court will benefit from deferring review until the agency’s policies have crystallized and the “question arises in some more concrete and final form.” 56 Should we decide that either the agency or the court has a significant interest in postponing review, we will decline to hear the petitioner’s claim at that time unless, under the “hardship to the parties” prong, the “interest of those who seek relief from the challenged action’s ‘immediate and practical impact’ upon them”57 outweighs the competing institutional interests in deferring review.58
[190]*190Applying the “fitness of the issues” prong to the instant case, we find initially that the issue presented for review— whether the HRS is arbitrary, capricious, or in excess of statutory authority — is a purely legal question and thus was “fit” in that respect for judicial resolution during the statutory period. We also find that neither the agency nor the court would have benefitted from postponing review beyond that period. Indeed, as we demonstrate below, both institutions had a significant interest in completing all review of the HRS during the statutory time frame.
The EPA’s interest in obtaining review of the HRS during the statutory period, before the model was applied to specific sites, may be demonstrated in three ways— by examining the statutory review provision itself, by considering the agency’s declared interest in the timing of review, and by taking .account of the EPA’s indications that this regulation represents its crystallized position on the HRS.
As noted earlier, CERCLA’s statutory review provision provides that
[rjeview of any regulation promulgated under [CERCLA] may be had upon application [to this court] by any interested person____ Any such application shall be made within ninety days from the date of promulgation of such regulations. Any matter with respect to which review could have been obtained under this [provision] shall not be subject to judicial review in any civil or criminal proceeding for enforcement or to obtain damages or recovery of response costs.59
It is abundantly clear from the plain language of this provision that Congress intended to provide prompt, uniform “preenforcement” review of CERCLA regulations to a broad class of petitioners, in order to avoid needless delays in the implementation of an important national program.60 This provision constitutes compelling evidence that Congress has, in effect, decided that the interest of the EPA in effectuating CERCLA’s purposes will generally be furthered by review during the statutory period and, by implication, hindered by postponing review. Indeed, Congress underscored its intention that review be expeditious by proscribing review at any stage of enforcement.61 In determining a regulation’s ripeness for review, courts should accord heavy weight to this sort of strong congressional proclamation of an agency’s interest in the timing of review of its regulations. This is particularly true in a case such as this, where the agency does not contend that it would benefit from deferral.
In fact, the EPA asserted at oral argument and in its brief that it would benefit if review of the legality of the HRS were confined to the statutory period. If, as the agency explained, the HRS could be challenged by different petitioners each time the NPL is updated, the EPA would be [191]*191forced, contrary to the will of Congress, to defend the HRS repeatedly, wasting both time and funds that would be better spent cleaning up hazardous wastes that threaten human health and the environment.62
A final consideration bearing on the agency’s interest in the timing of review is whether the HRS represented the agency’s “final position” on the methodology to be used in evaluating sites for inclusion on the NPL. As we observed in Continental Air Lines, Inc., v. CAB,
[t]he interest in postponing review is strong if the agency position whose validity is in issue is not in fact the agency’s final position. If the position is likely to be abandoned or modified before it is actually put into effect, then its review wastes the court’s time and interferes with the process by which the agency is attempting to reach a final decision.63
We find that, at the time of the HRS’s promulgation, there existed every indication that it constituted the agency’s “final position.” It was issued as a regulation, after notice-and-comment proceedings, and contains no equivocal or tentative language as to the EPA’s intention to employ it in evaluating sites for inclusion on the NPL. The EPA’s commitment to this model was also manifested by its use of an earlier version of the HRS to score sites for listing on the Interim Priority List (“IPL”).64
Given the strong congressional preference for timing of review of CERCLA regulations expressed in section 113(a), the likelihood that postponing review would interfere with the agency’s ability to fulfill its statutory responsibilities, and the fact that the HRS represented the agency’s “final position,” we find, without reservation, that the agency possessed no interest in deferring review until application of the HRS to produce the NPL; on the contrary, the EPA had a positive interest in completing review of the HRS during the statutory period.,
Furthermore, we find that the court had no compelling interest in postponing review until the NPL was issued. First, because the issue presented for review is purely a legal one, it was suitable for review at the time the HRS was issued. Second, since the HRS represented the agency’s final position, review would not have wasted the court’s time. Third, the court has an interest in conserving its own resources by resolving challenges to agency action during the statutory period, rather than stretching them out over an indefinite period of time. Fourth, as we can confirm through hindsight, the application of the HRS to produce the NPL would not have significantly advanced our ability to deal with the legal issue presented nor would it have aided us in its resolution.65 The issue is the same today as it was during the statutory period and our understanding of it has not been enhanced by the development of a more specific factual background. While we have sometimes chosen to defer review of a general standard 66 when we believed it beneficial to wait until the controversy had a more developed factual context,67 it is clear beyond [192]*192peradventure that the validity of a rule can be ripe for review whether or not it has actually been improperly applied and enforced in a concrete factual setting.68
We need not proceed in this case to the second prong of the Abbott Laboratories test. As we previously explained, the purpose of the “hardship to the parties” analysis is to ascertain if the harm that deferring review will cause the petitioners outweighs the benefits it will bring the agency and the court. Because we have determined that the EPA and, to a lesser extent, the court had a positive interest in review during the statutory period, there are no conflicting interests to balance. The interests of the EPA, the court, and the petitioners in immediate review of the HRS would have coincided. Thus the ripeness doctrine would not have barred our review of the HRS within the statutory period.69
In some of our decisions we have suggested that the court should consider “the hardship to the parties,” even where the first prong of the Abbott Laboratories test is met.70 In none of these cases, however, did we explicitly find that both the agency and the court had a positive interest in immediate review. Nor did we consider the relevance to our ripeness determination of a statutory review provision that expresses a strong congressional preference concerning the timing of review. Where the first prong of the ripeness test is met and Congress has emphatically declared a preference for immediate review, assuming that constitutional case or controversy requirements have been met, no purpose is served by proceeding to the second prong.
Indeed, the mechanical application of the second prong of the Abbott Laboratories test could work mischief in such a situation. If we were to defer review in such a case merely because we could find no significant harm to the petitioner from delay, we would achieve the perverse result of postponing review to the detriment of the agency and the court and in contravention of the express preferences of Congress, in the name of a prudential doctrine that is intended to protect the institutional needs of courts and agencies. As we have repeatedly observed, the ripeness doctrine requires a pragmatic and commonsense appli[193]*193cation.71 Common sense, and the underlying goals of the ripeness doctrine, are both furthered by finding that the HRS was ripe for review during the statutory period.
We conclude that the petitioners were not justified in failing to seek review during the statutory period. We recognize, however, that the relationship between the ripeness doctrine and statutory review provisions that exhibit a strong congressional preference for “pre-enforcement” review was largely uncharted before this decision. For this reason, we proceed to consider the merits of the petitioners’ challenge.
III. The Petitioners’ Challenge to the HRS
The petitioners attack the HRS on two grounds. First, they argue that it is inconsistent with CERCLA’s purposes. Second, they contend that it is arbitrary, capricious and an abuse of discretion.
A. Argument that the HRS is Inconsistent with CERCLA’s Purposes
CERCLA clearly contemplates that the EPA’s limited funds and cleanup energies will be spent on releases or potential releases of harmful substances that threaten serious harm and that the EPA will set priorities for taking action with regard to these releases.72 Essentially, the petitioners argue that these statutory goals cannot be realized unless CERCLA is read to mean that the 400 sites listed on the first NPL must represent a definitive statement of the most hazardous sites in the country, at which EPA action may immediately be taken. Because the HRS is designed to ascertain merely if a site poses a threat to human health or to the environment that warrants further investigation, but is not sufficiently refined to make final determinations of which sites pose threats that definitely qualify for action under CERCLA, the petitioners maintain that it produces a NPL that is inconsistent with CERCLA’s purposes.
The EPA argues that nothing in the statute requires so narrow a view of the HRS and NPL. The agency interprets CERCLA to permit it to establish a lower minimum level of certainty that a release or potential release poses a threat for purposes of listing on the NPL than for actually taking government action. The EPA considers the HRS to be a useful tool for sifting through a large number of sites in a relatively expeditious and inexpensive manner to pinpoint problem sites which deserve more comprehensive, and thus costly, analysis. In the preamble to the final rule promulgating the HRS, the agency noted, in response to comments similar to the complaints lodged by the petitioners here, that the
role and importance of the HRS and NPL must be kept in perspective, the [sic] NPL, which will include at least 400 releases, is merely the first step in considering a release for Fund-financed remedial response. If a release is included on the NPL but a later remedial investigation discloses the hazard to be less significant than originally thought to be, a decision may be made not to provide Fund-financed remedial response. Similarly, the NPL will be reviewed periodically and a release can be added if more extensive data indicate a more significant hazard at the release.73
The agency elaborated on this view of the HRS and NPL in the preamble to the final version of the NPL:
The purpose of the NPL ... is primarily to serve as an informational tool for use by EPA in identifying sites that appear to present a significant risk to public health or the environment. The initial identification of a site in the NPL is intended primarily to guide EPA in determining which sites warrant further investigation designed to assess the nature and extent of the public health and envi[194]*194ronmental risks associated with the site and to determine what response action, if any, may be appropriate. Inclusion of a site on the NPL does not establish that EPA necessarily will undertake response actions. Moreover, listing does not require any action of any private party, nor does it determine the liability of any party for the cost of cleanup at the site.
In addition, although the HRS scores used to place sites on the NPL may be helpful to the Agency in determining priorities for cleanup and other response activities among sites on the NPL, EPA does not rely on the scores as the sole means of determining such' priorities, as discussed below. Neither can the HRS itself determine the appropriate remedy for a site. The information collected to develop HRS scores to choose sites for the NPL is not sufficient in itself to. determine the appropriate remedy for a particular site. After a site has been included on the NPL, EPA generally will rely on further, more detailed studies conducted at the site to determine what response, if any, is appropriate____ After conducting these additional studies EPA may conclude that it is not feasible to conduct response action at some sites on the NPL because of more pressing needs at other sites. Given the limited resources available in the Hazardous Substance Response Fund, the Agency must carefully balance the relative needs for response at the numerous sites it has studied. It is also possible that EPA will conclude after further analysis that no action is needed at the site because the site does not present a problem.74
We are persuaded that the EPA’s interpretation of CERCLA is a reasonable one which should be upheld. Our decision is guided by the Supreme Court’s opinion from last Term, Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc.75 In Chevron, the Court explained that when “a court reviews an agency’s construction of the statute which it administers, ... if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.”76 If the agency’s “ ‘choice represents a reasonable accommodation of conflicting policies that were committed to [its] care by the statute, [a court] should not disturb it unless it appears from the statute or its legislative history that the accommodation is not one that Congress would have sanctioned.’ ” 77
Applying these principles here, we see that the EPA has been entrusted with the administration of CERCLA. Nothing in the language of the statute explicitly speaks to the question whether the standards for listing a site on the NPL and for taking response action must be identical. Furthermore, CERCLA’s legislative history firmly supports the agency’s position. The Senate Report recognized that the NPL would have to be “based on information immediately available” to the EPA,78 and went on to observe that
[t]he priority lists serve primarily informational purposes, identifying for the States and the public those facilities and sites or other releases which appear to warrant remedial action. Inclusion of a facility or site on the list does not in itself reflect a judgment of the activities of its owner or operator, it does not require those persons to undertake any action, nor does it assign liability to any person. Subsequent government action in the form of remedial actions or enforcement action will be necessary in order to do so, and these actions will be [195]*195attended by all appropriate procedural safeguards.79
Clearly, the EPA’s decision to reconcile the need for certainty before action with the need for inexpensive, expeditious procedures to identify potentially hazardous sites by establishing different threshold criteria for action and for listing is reasonable and fully in accord with congressional intent. We uphold it here.
B. Argument that the HRS is Arbitrary, Capricious or an Abuse of Discretion
The petitioners also question the rationality of the HRS on several grounds, primarily in regard to its ability to evaluate mining waste sites. The petitioners contend that the HRS was designed with the paradigm of a chemical waste site in mind, and that it consequently incorporates certain assumptions which, although pertinent to chemical dump sites, have no relevance to mining waste sites. In particular, they allege that: (1) the basis for scoring “observed releases” is unreasonable, basically because that score reflects only the fact of release and not the severity of the release; (2) the basis for scoring waste characteristics is unreasonable because it fails to consider the relatively low concentration of harmful substances present in high-volume mining wastes; (3) the basis for scoring the “target” population threatened by a release is unreasonable because it is rooted in a formula for estimating the population within a certain radius of the release and does not utilize actual population figures; and (4) the HRS should take into account ongoing or completed remedial measures.
The EPA responds to the petitioners’ argument in detail. In essence, the agency contends that the HRS is a useful and reasonable tool that accomplishes its objective — the rough estimation of which sites, containing either chemical or mining wastes, deserve further attention.
The standard of review applicable here is that set forth in section 10(e)(2)(A) of the Administrative Procedure Act, whether the EPA's action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”80 Under the arbitrary and capricious standard we look to see if the agency has examined relevant data and has articulated a rational explanation for its action.81 After considering the arguments and thoroughly reviewing the record, we find that neither the HRS itself, nor its use to evaluate mining waste sites, violates the foregoing standard.
An agency may utilize a predictive model so long as it explains the assumptions and methodology it used in preparing the model.82 If the model is challenged, the agency must provide a full analytical defense.83 But, as this court recently held in Small Refiner Lead Phase-Down Task Force v. EPA, while a reviewing court
will also look for evidence that the agency is conscious of the limits of the model ... [ultimately, however, we must defer to the agency’s decision on how to balance the cost and complexity of a more elaborate model against the oversimplification of a simpler model. We can reverse only if the model is so oversimpli[196]*196fied that the agency’s conclusions from it are unreasonable.84
In the preambles to the final versions of the HRS85 and the NPL,86 the EPA offered thorough and reasoned explanations of the assumptions and methodology it relied upon in creating the HRS. Throughout these preambles, the agency clearly indicated its awareness of the limitations of the model,87 including those regarding its application to mining waste sites.88 This awareness is further manifested in the EPA’s decision to use the HRS only to make a preliminary division between sites that justify further consideration and those that do not, rather than to employ it to decide which sites warrant response action under CERCLA. Moreover, the EPA adequately addressed the substance of each of the petitioners’ complaints in its response to comments.89 We find that the agency’s explanations are reasonable and see no reason to rehearse them here.
IV Conclusion
The petitioners’ challenge to the HRS is an untimely claim that was ripe for review during the statutory period, and thus would normally be time-barred. Because, however, in this opinion we clarify for the first time certain aspects of the relationship between statutory review provisions and the ripeness doctrine, we offer an alternative holding on the merits of the petitioners’ claim. Given the relatively limited, informational purposes of the HRS and the NPL, we find the HRS itself, and its use in evaluating mining waste sites, to be reasonable and consistent with the purposes of the statute. The petitioners’ request for relief is
Denied.