with whom
Justice Thomas and Justice Breyer join, dissenting.The Court today disregards the Federal Rules of Civil Procedure (Civil Rules) in habeas corpus cases, chiefly because it believes that this departure will make no difference. See ante, at 209. Even if that were true, which it is not, I could not join this novel presumption against applying the Civil Rules.
The Civil Rules “govern the procedure in the United States district courts in all suits of a civil nature.” Rule 1. This includes “proceedings for ... habeas corpus,” Rule 81(a)(2), but only “to the extent that the practice in such proceedings is not set forth in statutes of the United States [or] the Rules Governing Section 2254 Cases” (Habeas Rules), Civil Rule 81(a)(2); see also Habeas Rule 11. Thus, “[t]he Federal Rules of Civil Procedure apply in the context of habeas suits to the extent that they are not inconsistent with the Habeas Corpus Rules,” Woodford v. Garceau, 538 U. S. 202, 208 (2003), and do not contradict or undermine the provisions of the habeas corpus statute, Gonzalez v. Crosby, 545 U. S. 524, 529-530 (2005).
As the Court notes, the Civil Rules adopt the traditional forfeiture rule for unpleaded limitations defenses. See ante, at 207-208 (citing Rules 8(c), 12(b), 15(a)). The Court does not identify any “inconsisten[cy]” between this forfeiture rule and the statute, Rules, or historical practice of habeas proceedings — because there is none. Forfeiture of the limitations defense is demonstrably not inconsistent with traditional habeas practice, because, as the Court acknowledges, habeas practice included no statute of limitations until 1996. Ante, at 202, n. 1; see also infra, at 214-216. Forfeiture is perfectly consistent with Habeas Rule 5(b), which now provides that the State’s “answer . . . must state whether any *213claim in the petition is barred by ... statute of limitations.” (Emphasis added.) And forfeiture is also consistent with (and indeed, arguably suggested by) Habeas Rule 4, because Rule 4 provides for sua sponte screening and dismissal of habeas petitions only prior to the filing of the State’s responsive pleading.1
Most importantly, applying the forfeiture rule to the limitations period of 28 U. S. C. § 2244(d) does not contradict or undermine any provision of the habeas statute. Quite the contrary, on its most natural reading, the statute calls for the forfeiture rule. AEDPA expressly enacted, without further qualification, “[a] 1-year period of limitation” for habeas applications by persons in custody pursuant to the judgments of state courts. § 2244(d)(1) (emphasis added). We have repeatedly stated that the enactment of time-limitation periods such as that in § 2244(d), without further elaboration, produces defenses that are nonjurisdictional and thus subject to waiver and forfeiture. See Zipes v. Trans World Airlines, Inc., 455 U. S. 385, 393 (1982); see also Eberhart v. United States, 546 U. S. 12, 15 (2005) (per curiam); Kontrick v. Ryan, 540 U. S. 443, 447 (2004). Absent some affirmative incompatibility with habeas practice, there is no reason why a habeas limitations period should be any different. By imposing an unqualified “period of limitation” against the background understanding that a defense of “limitations” must be raised in the answer, see Civil Rules 8(c), 12(b), the statute implies that the usual forfeiture rule is applicable.
*214Instead of identifying an inconsistency between habeas corpus practice and the usual civil forfeiture rule, the Court urges that “it would make scant sense to distinguish in this regard AEDPA’s time bar from other threshold constraints on federal habeas petitioners” that may be raised sua sponte — ante, at 209 — namely, exhaustion of state remedies, procedural default, nonretroactivity, and (prior to AEDPA) abuse of the writ. See Granberry v. Greer, 481 U. S. 129, 133 (1987) (exhaustion); Caspari v. Bohlen, 510 U. S. 383, 389 (1994) (nonretroaetivity). But unlike AEDPA’s statute of limitations, these defenses were all created by the habeas courts themselves, in the exercise of their traditional equitable discretion, see Withrow v. Williams, 507 U. S. 680, 717-718 (1993) (Scalia, J., concurring in part and dissenting in part), because they were seen as necessary to protect the interests of comity and finality that federal collateral review of state criminal proceedings necessarily implicates. See McCleskey v. Zant, 499 U. S. 467, 489-491 (1991) (abuse of the writ); Wainwright v. Sykes, 433 U. S. 72, 80-81 (1977) (procedural default); Teague v. Lane, 489 U. S. 288, 308 (1989) (nonretroactivity); Rose v. Lundy, 455 U. S. 509, 515 (1982) (exhaustion of state remedies). Unlike these other defenses, no time limitation — not even equitable laches — was imposed to vindicate comity and finality. AEDPA’s 1-year limitations period is entirely a recent creature of statute. See ante, at 202, n. 1. If comity and finality did not compel any time limitation at all, it follows a fortiori that they do not compel making a legislatively created, forfeitable time limitation nonforfeitable.
In fact, prior to the enactment of AEDPA, we affirmatively rejected the notion that habeas courts’ traditionally broad discretionary powers would support their imposition of a time bar. Historically, “there [wa]s no statute of limitations governing federal habeas, and the only laches recognized [wa]s that which affects the State’s ability to defend against the claims raised on habeas” — which was imposed by *215Rule, and not until 1977. Brecht v. Abrahamson, 507 U. S. 619, 637 (1993); see also United States v. Smith, 331 U. S. 469, 475 (1947); 17A C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §4268.2, pp. 497-498 (2d ed. 1988) (hereinafter Wright & Miller). We repeatedly asserted that the passage of time alone could not extinguish the habeas corpus rights of a person subject to unconstitutional incarceration. See Pennsylvania ex rel. Herman v. Claudy, 350 U. S. 116, 123 (1956); Chessman v. Teets, 354 U. S. 156, 164-165 (1957). For better or for worse, this doctrine was so well entrenched that the lower courts regularly entertained petitions filed after even extraordinary delays. See, e. g., Hawkins v. Bennett, 423 F. 2d 948, 949 (CA8 1970) (40 years); Hamilton v. Watkins, 436 F. 2d 1323, 1325 (CA5 1970) (at least 36 years); Hannon v. Maschner, 845 F. 2d 1553, 1553-1555 (CA10 1988) (at least 24 years). And in 1977, when enactment of the former Habeas Rule 9(a) “introduce [d] for the first time an element of laches into habeas corpus,” 17A Wright & Miller § 4268.2, at 498 — by adopting the rule against “‘prejudicial delay’” to which the Court refers, ante, at 202, n. 1 — even that limited doctrine was treated as subject to the very same pleading requirements and forfeiture rule that the Court rejects today for the stricter limitations period of § 2244(d). See Smith v. Secretary of New Mexico Dept. of Corrections, 50 F. 3d 801, 821-822, n. 30 (CA10 1995); see also McDonnell v. Estelle, 666 F. 2d 246, 249 (CA5 1982).
There is, therefore, no support for the notion that the traditional equitable discretion that governed habeas proceedings permitted the dismissal of habeas petitions on the sole ground of untimeliness. Whether or not it should have, see Collins v. Byrd, 510 U. S. 1185, 1186-1187 (1994) (SCALIA, J., dissenting), it did not. The Court’s reliance on pre-existing equitable doctrines like procedural default and nonretroactivity is, therefore, utterly misplaced. Nothing in our tradition of refusing to dismiss habeas petitions as untimely *216justifies the Court’s decision to beef up the presumptively forfeitable “limitations period” of § 2244(d) by making it the subject of sua sponte dismissal.
In what appears to be the chief ground of its decision, the Court also observes that “the Magistrate Judge, instead of acting sua sponte, might have informed the State of its obvious computation error and entertained an amendment to the State’s answer” under Civil Rule 15(a). Ante, at 209. Although “an amendment to the State’s answer might have obviated this controversy,” the Court concedes, “we see no dis-positive difference between that route, and the one taken here.” Ibid. But this consideration cuts in the opposite direction. If there truly were no “dispositive difference” between following and disregarding the rules that Congress has enacted, the natural conclusion would be that there is no compelling reason to disregard the Civil Rules.2 Legislatively enacted rules are surely entitled to more respect than this apparent presumption that, when nothing substantial hangs on the point, they do not apply as written. And, unlike the novel regime that the Court adopts today, which will apparently require the development of new rules from scratch, there already exists a well-developed body of law to govern the district courts’ exercise of discretion under Rule *21715(a). See 6 Wright & Miller §§ 1484-1488 (2d ed. 1990 and Supp. 2005). Ockham is offended by today’s decision, even if no one else is.
But, in fact, there are at least two notable differences between the Civil Rules and the sua sponte regime of such cases as Granberry and Caspari — both of which involve sufficiently significant departures from ordinary civil practice as to require clear authorization from the statute, the Rules, or historical habeas practice. First, the Granberry regime allows the forfeited procedural defense to be raised for the first time on appeal, either by the State or by the appellate court sua sponte. See 481 U. S., at 130, 133; Schiro v. Farley, 510 U. S. 222, 228-229 (1994). Ordinary civil practice does not allow a forfeited affirmative defense whose underlying facts were not developed below to be raised for the first time on appeal. See Weinberger v. Salfi, 422 U. S. 749, 764 (1975); Metropolitan Housing Development Corp. v. Arlington Heights, 558 F. 2d 1283, 1287 (CA7 1977). The ability to raise even constitutional errors in criminal trials for the first time on appeal is narrowly circumscribed. See Fed. Rule Crim. Proc. 52(b); United States v. Olano, 507 U. S. 725, 732 (1993). Comity and finality justified this departure from ordinary practice for historically rooted equitable defenses such as exhaustion. See Granberry, supra, at 134. But limitations was not such a defense.
Also, Granberry and the like raise the possibility that the courts can impose a procedural defense over the State’s affirmative decision to waive that defense. The Court takes care to point out that this is not such a case, ante, at 210-211, but it invites such cases in the future. After all, the principal justification for allowing such defenses to be raised sua sponte is that they “ ‘implicat[e] values beyond the concerns of the parties,’ ” including “ ‘judicial efficiency and conservation of judicial resources’ ” and “the expeditious handling of habeas proceedings.” Ante, at 205, 208 (quoting Acosta v. Artuz, 221 F. 3d 117, 123 (CA2 2000)). There are *218many reasons why the State may wish to disregard the statute of limitations, including the simple belief that it would be unfair to impose the limitations defense on a particular defendant. On the Court’s reasoning, a district court would not abuse its discretion in overriding the State’s conscious waiver of the defense in order to protect such “ Values beyond the concerns of the parties,’ ” ante, at 205.3 Under the Civil Rules, by contrast, amending a party’s pleading over his objection would constitute a clear abuse of the trial court’s discretion.
In sum, applying the ordinary rule of forfeiture to the AEDPA statute of limitations creates no inconsistency with the Habeas Rules. On the contrary, it is the Court’s unwar*219ranted expansion of the timeliness rule enacted by Congress that is inconsistent with the statute, the Habeas Rules, the Civil Rules, and traditional practice. I would hold that the ordinary forfeiture rule, as codified in the Civil Rules, applies to the limitations period of § 2244(d). I respectfully dissent.
The Court observes that “[w]ere we to accept Day’s position, courts would never (or, at least, hardly ever) be positioned to raise AEDPA’s [Antiterrorism and Effective Death Penalty Act of 1996] time bar sua sponte,” because “information essential to the time calculation is often absent” at the Rule 4 prescreening stage, ante, at 207, n. 6. But to be distressed at this phenomenon is to beg the question — that is, to assume that courts ought to “be positioned to raise AEDPA’s time bar sua sponte.” That is precisely the question before us.
I agree with the Court that today’s decision will have little impact on the outcome of district court proceedings. In particular, I agree that “if a [district] judge does detect a clear computation error, no Rule, statute, or constitutional provision commands the judge to suppress that knowledge,” ante, at 210. Rather, a judge may call the timeliness issue to the State’s attention and invite a motion to amend the pleadings under Civil Rule 15(a), under which “leave shall be freely given when justice so requires.” In fact, in providing for leave whenever “justice so requires,” Rule 15(a), the Civil Rules fully accommodate the comity and finality interests that the Court thinks require a departure from the Civil Rules, see ante, at 206, 210. Requiring the State to take the affirmative step of amending its own pleading at least observes the formalities of our adversary system, which is a nontrivial value in itself. See United States v. Burke, 504 U. S. 229, 246 (1992) (Scalia, J., concurring in judgment).
In order to avoid this seemingly unavoidable conclusion, the Court asserts, without relevant citation or reasoning, that “should a State intelligently choose to waive a statute of limitations defense, a district court would not be at liberty to disregard that choice.” Ante, at 210, n. 11. This assertion is contrary to our statement in Granberry v. Greer, 481 U. S. 129, 134 (1987) — a case which, on the Court’s view, it makes “scant sense to distinguish,” ante, at 209 — that an appellate court may dismiss an unexhausted petition sua sponte in “cases in which the State fails, whether inadvertently or otherwise, to raise an arguably meritorious non-exhaustion defense.” (Emphasis added.) To support its assertion, the Court cites nothing but its own earlier statement: “Ordinarily in civil litigation, a statutory time limitation is forfeited if not raised in a defendant’s answer or in an amendment thereto. Fed. Rules Civ. Proc. 8(c), 12(b), and 15(a). And we would count it an abuse of discretion to override a State’s deliberate waiver of a limitations defense.” Ante, at 202. But as the statement itself shows, the “ordinary” inability to override the State’s “intelligent” waiver is coupled with an “ordinary” automatic forfeit of the defense if it is not timely raised. The Court does not say why it makes sense, for the statute of limitations of § 2244(d)(1)(A), to reject (as it does) the first part of the ordinary practice (automatic forfeiture), while embracing the second (inability to override intelligent waiver). The reason for rejecting the first part surely applies just as well to the second: Section 2244(d)(1)(A) supposedly “ ‘implicate^] values beyond the concerns of the parties,’ ” including “ ‘judicial efficiency,’ ” “ ‘conservation of judicial resources,’” and “expeditious handling of habeas proceedings.” Ante, at 205, 208.