with whom Justice Scalia, Justice Kennedy, and Justice Thomas join, concurring in the judgment.
I agree with the Court that the judgment of the Court of Appeals should be reversed, but I am in sufficient disagreement with the Court’s reasoning to write separately. I disagree with the Court’s statement that “the Court of Appeals order vacating the stay is lawful only if dismissal of the petition would have been lawful.” Ante, at 319. This statement, I believe, misreads our opinion in Barefoot v. Estelle, 463 U. S. 880 (1983), and ignores our reasoning in Gomez v. United States Dist. Court for Northern Dish of Cal., 503 U. S. 653 (1992).
The order under review does not dispose of a petition for a writ of habeas corpus; it vacates a stay of execution. The Court dismisses this distinction as a “preliminary matter,” which “makes no difference,” because “the Court of Appeals order vacating the stay is lawful only if dismissal of the petition would have been lawful.” Ante, at 319. In my view, *335the fact that we are reviewing an order vacating a stay is anything but “preliminary.”
The Court is correct inasmuch as the underlying petition’s likelihood of success is one factor to be considered in determining whether a stay should be entered. See Hilton v. Braunskill, 481 U. S. 770, 776 (1987). Rule 9 of the Rules Governing § 2254 Cases sets forth the grounds upon which a habeas petition may be dismissed other than the merits. Under Rule 9(b), a petition may be dismissed if it is found to be successive or abusive. Under Rule 9(a) it may also be dismissed
“if it appears that the state of which the respondent is an officer has been prejudiced in its ability to respond to the petition by delay in its filing unless the petitioner shows that it is based on grounds of which he could not have had knowledge by the exercise of reasonable diligence before the circumstances prejudicial to the state occurred.” 28 U. S. C. § 2254 Rule 9(a).
In this case, there was no basis for denying a stay on the ground that petitioner’s habeas claims are without merit; for the purposes of considering the stay application, it is undisputed that those claims are substantial. Because the habeas petition was petitioner’s first, it would also have been inappropriate to deny a stay on the ground that the petition could have been dismissed under Rule 9(b). I agree with the majority, ante, at 326, that, on the record before us, the petition likewise could not have been dismissed under Rule 9(a), because the Rule’s elements were not satisfied. Although the District Court determined that petitioner engaged in delay, it made no determination that the delay prejudiced the State’s ability to respond to the petition, within the meaning of Rule 9(a), by depriving the State of adequate time to respond or otherwise.
However, an applicant’s likelihood of success is not the only consideration in determining whether he is entitled to a stay. *336See Hilton, supra, at 776. The Habeas Rules say little if anything about this determination, and understandably so. It must be remembered the statutes governing habeas corpus, 28 U. S. C. §§2241-2255, were enacted in 1948, before the writ developed into a vehicle for federal courts “to reexamine federal constitutional issues even after trial and review by a state,” Brown v. Allen, 344 U. S. 443, 459 (1953), and long before this Court declined to declare the death penalty unconstitutional, in Gregg v. Georgia, 428 U. S. 153 (1976), Proffitt v. Florida, 428 U. S. 242 (1976), and Jurek v. Texas, 428 U. S. 262 (1976). It would have been difficult for Congress to have anticipated the issues that arise in a system in which state death sentences are presumptively valid, but are “reexamined” in federal court before execution to consider constitutional challenges to the manner in which they were imposed.
In the typical noncapital habeas case, it is relatively easy to rule on an application to stay execution of a state sentence by consulting ordinary principles governing stays. Rarely, if ever, does a noncapital petitioner seek a stay of his sentence before the district court has passed on the merits of his petition. When a petitioner does make such a request, he usually has little chance of success on the merits, since he has been confined pursuant to a presumptively valid final judgment of conviction rendered by a state court. See 28 U. S. C. § 2254(d). If, after entertaining his petition, the district court awards the writ, the “stay equities” shift in favor of the petitioner, who will be enlarged unless the State can demonstrate that the equities counsel otherwise. Hilton, supra, at 774.
This easily managed system can be adapted to govern capital habeas cases, so long as the capital petitioner files his habeas petition sufficiently in advance of his execution date. If he files in a timely fashion, the district court may then consider the petition in due course, without in any way disturbing the sentence or execution date before ruling on the *337petition’s merits. But if, as in this case, the petitioner instead flies an “eleventh hour” federal habeas petition, the customary principles must be revised accordingly. The district court may feel that it simply does not have time before the date of execution to adequately consider the merits of petitioner’s claims, and will naturally be disposed, as the District Court was here, to enter a stay to enable it to do so.1 In so doing, the district court sets aside a scheduled state execution of sentence, imposed by a presumptively valid final state judgment of conviction, on the basis of a tentative assessment that the judgment violates a federal constitutional right. Unless the eleventh-hour nature of the petition is taken into account, the late filing may induce the federal court to disregard federal-state comity and “frustrate ... the States’ sovereign power to punish offenders,” Engle v. Isaac, 456 U. S. 107, 128 (1982), when such interference might have been avoided by timely filing.2 The customary principles must also be revised to account for an attempt by a petitioner to manipulate the district court into granting relief where relief is clearly precluded.
In Gomez v. United States Dist. Court for Northern Dist. of Cal., 503 U. S. 653 (1992), this Court demonstrated how last-minute or manipulative uses of the stay power constitute equitable grounds which can justify the denial of an application for stay of a state-court order of execution. The Court *338vacated a stay of execution issued on behalf of Robert Alton Harris, a California prisoner, pending consideration of a 42 U. S. C. § 1983 action alleging that his method of execution violated the Eighth Amendment. See 503 U. S., at 653. Because Harris had not raised the Eighth Amendment claim in any of the four federal habeas corpus petitions he had filed over 10 years, the Court considered the § 1983 claim “an obvious attempt to avoid the application of McCleskey v. Zant, 499 U. S. 467 (1991), to bar this successive claim for relief.” Ibid. We could have vacated the stay on the basis of the successive-petition bar alone, but we explicitly did not:
“Even if we were to assume, however, that Harris could avoid the application of McCleskey to bar his claim, we would not consider it on the merits. Whether his claim is framed as a habeas petition or as a § 1983 action, Harris seeks an equitable remedy. Equity must take into consideration the State’s strong interest in proceeding with its judgment and Harris’ obvious attempt at manipulation. This claim could have been brought more than a decade ago. There is no good reason for this abusive delay, which has been compounded by last-minute attempts to manipulate the judicial process. A court may consider the last-minute nature of an application to stay execution in deciding whether to grant equitable relief.” Id., at 653-654 (citations omitted).
Our order confirms that “abusive delay” — waiting until the last minute to submit a claim that could have been submitted earlier — and “obvious attempts] at manipulation” — in that case, asking the court to exercise its equitable powers in defiance of a clearly applicable legal rule precluding relief on the merits — constitute equities to be considered in ruling on the prayer for relief. More important, because we explained that this misconduct constituted sufficient grounds to deny Harris’ stay application, “[ejven if” McCleskey did not bar *339his claim, we made clear that such abuse may tip the scales decisively against a stay applicant regardless of the applicant’s likelihood of success on the merits.3
Gomez also confirms that a habeas petitioner’s misconduct in applying for a stay may disentitle him to the stay even if the petition is his first. The inequitable conduct Gomez criticized, abusive delay and manipulation, may be present in any stay application. In Gomez we did not equivocate when we said: “Equity must take into consideration [an] obvious attempt at manipulation. ... A court may consider the last-minute nature of an application to stay execution in deciding whether to grant equitable relief.” 503 U. S., at 654 (citations omitted). It may be admitted that there would be a stronger presumption in favor of deciding the merits of a first-time petition than for a successive petition. The successive nature of a petition gives rise to an additional concern counseling against review of the merits: that the petitioner is frustrating the State’s- attempts to execute its judgment by exploiting, the fact that ordinary principles of res judicata do not apply in habeas corpus. McCleskey v. Zant, 499 U. S. 467, 479, 491-492 (1991). It does not follow, however, that because a first-time petitioner does not abuse the writ under Rule 9(b) he may never be found to have engaged in the misconduct we criticized in Gomez; it means only that misconduct by such a petitioner is less likely to result in a refusal to grant a stay in order to consider the merits of the petition.
The majority attempts to distinguish Gomez because the matter before the Court in that case was “not a first habeas petition.” Ante, at 321. This reading is wholly implausible, because the first paragraph of the Court’s order had already discussed the fact that Harris was not a first-time petitioner. *340If the paragraph about Harris’ misconduct in relation to his application, quoted above, had legal significance only if his petition was successive, it would have been superfluous.
To support its view that a stay must be granted if a first federal habeas petition is not dismissed, the Court relies on our decision in Barefoot v. Estelle, 463 U. S. 880 (1983). But Barefoot and the present case arose in different contexts. The question presented and decided in Barefoot only addressed how the merits of the habeas petition may determine whether the petitioner obtains a stay. Id., at 887 (announcing the Court was considering “the appropriate standard for granting or denying a stay of execution pending disposition of an appeal by a federal court of appeals by a death-sentenced federal habeas corpus petitioner”); id., at 891 (affirming the denial of a stay because the Court of Appeals “ruled on the merits of [Barefoot’s] appeal”). The issue in the present case is quite different: whether a petitioner’s course of conduct in seeking the writ may be considered by the district court in deciding whether to grant a stay. To the extent that the Court’s reading of Barefoot depends on the belief that a decision on a first federal habeas petition is somehow necessary to validate a state conviction, the Court ignores Barefoot’s assertion to the contrary:
“The role of federal habeas proceedings ... is secondary and limited. Federal courts are not forums in which to relitigate state trials. . . . The procedures adopted to facilitate the orderly consideration and disposition of ha-beas petitions are not legal entitlements that a defendant has a right to pursue irrespective of the contribution these procedures make toward uncovering constitutional error.” Id., at 887-888.
And, contrary to the Court’s refusal to consider whether Barefoot’s “rationale” might brook a distinction between seasonable and eleventh-hour first habeas petitions, ante, at 321, our opinion warned that federal habeas corpus is not “a *341means by which a defendant is entitled to delay an execution indefinitely.” 463 U. S., at 887.
I nonetheless agree with the Court that the Court of Appeals erred in vacating the stay granted in this case by the District Court. The District Court did not consider whether petitioner’s conduct in court constituted misconduct so abusive that it disentitled him to a stay; it focused solely on the likelihood that petitioner’s habeas petition might be dismissed. Although the court determined that petitioner had “abused the writ,” it did not rely on this finding to deny a stay, correctly concluding that a first habeas petition may not be dismissed on the basis of abuse of the writ. App. 61-62. There was no determination that petitioner’s habeas petition could be dismissed under Rule 9(a). There is no other ground under which to dismiss a first petition other than the merits, and the Court of Appeals erred in concluding otherwise.
Although the findings supporting the District Court’s determination that petitioner abused the writ would go a long way toward supporting affirmance on the ground that petitioner’s misconduct disentitled him to a stay, reversal is still in order. I agree with the Court that petitioner’s conduct in the next-friend proceedings “neither aggravate nor mitigate Lonchar’s delay in filing.” Ante, at 331. Petitioner may not be blamed for having asserted his competence and his control over his habeas claims, because our case law required the District Court to establish as much. See Whitmore v. Arkansas, 495 U. S. 149, 165 (1990). Nor should he be blamed for his brother’s and sister’s desire to protect him, although it would be a different case if the record established that his relatives and he were colluding to stay his execution but avoid putting his claims before the court, so as to keep his options open in the future. Because the District Court erred in concluding that petitioner was culpable for the course of the next-friend proceedings and “[i]t is a paradigmatic abuse of discretion for a court to base its judgment on *342an erroneous view of the law,” Schlup v. Delo, 513 U. S. 298, 333 (1995) (O’Connor, J., concurring) (citing Cooter & Gell v. Hartmarx Corp., 496 U. S. 384, 405 (1990)), the District Court necessarily abused its discretion in determining there was abuse of the writ. Reconsideration of this determination and the other equities of petitioner’s stay application is now in order. I therefore concur in the judgment of reversal.
The Court is not concerned by this prospect because district courts have discretion to “order expansion of the record,” authorize discovery, decide “whether to hold an evidentiary hearing,” and generally “expedite proceedings.” Ante, at 326. These tools are useless, however, when a petitioner deliberately leaves the district court only one day to review a petition’s claims.
Of course, there may be cases in which the eleventh-hour nature of the petition is attributable to the State’s scheduling the execution date before the petitioner may appeal the denial of postconviction relief in a timely manner, not to the petitioner’s deliberate refusal to seek relief. I am certain that district courts are capable of distinguishing between the two situations.
The § 1983 action was a class-action suit, of which Harris was one member. I note that the claim that the Court declined to consider on the merits has enough merit for the class plaintiffs to have prevailed in district court and the court of appeals. Fierro v. Gomez, 77 F. 3d 301 (CA9 1996).