with whom The Chief Justice and Justice Thomas join, dissenting.
The Court today holds that the habeas statute, 28 U. S. C. § 2241, extends to aliens detained by the United States military overseas, outside the sovereign borders of the United States and beyond the territorial jurisdictions of all its courts. This is not only a novel holding; it contradicts a half-century-old precedent on which the military undoubtedly relied, Johnson v. Eisentrager, 339 U. S. 763 (1950). The Court’s contention that Eisentrager was somehow ne*489gated by Braden v. 30th Judicial Circuit Court of Ky., 410 U. S. 484 (1973) — a decision that dealt with a different issue and did not so much as mention Eisentrager — is implausible in the extreme. This is an irresponsible overturning of settled law in a matter of extreme importance to our forces currently in the field. I would leave it to Congress to change §2241, and dissent from the Court’s unprecedented holding.
I
As we have repeatedly said: “Federal courts are courts of limited jurisdiction. They possess only that power authorized by Constitution and statute, which is not to be expanded by judicial decree. It is to be presumed that a cause lies outside this limited jurisdiction ....” Kokkonen v. Guardian Life Ins. Co. of America, 511 U. S. 375, 377 (1994) (citations omitted). The petitioners do not argue that the Constitution independently requires jurisdiction here.1 Accordingly, these cases turn on the words of § 2241, a text the Court today largely ignores. Even a cursory reading of the habeas statute shows that it presupposes a federal district court with territorial jurisdiction over the detainee. Section 2241(a) states:
“Writs of habeas corpus may be granted by the Supreme Court, any justice thereof, the district courts and any circuit judge within their respective jurisdictions.” (Emphasis added.)
It further requires that “[t]he order of a circuit judge shall be entered in the records of the district court of the district wherein the restraint complained of is had.” (Emphases added.) And §2242 provides that a petition “addressed to the Supreme Court, a justice thereof or a circuit judge . . . *490shall state the reasons for not making application to the district court of the district in which the-applicant is held” (Emphases added.) No matter to whom the writ is directed, custodian or detainee, the statute could not be clearer that a necessary requirement for issuing the writ is that some federal district court have territorial jurisdiction over the detainee. Here, as the Court allows, see ante, at 478, the Guantanamo Bay detainees are not located within the territorial jurisdiction of any federal district court. One would think that is the end of these cases.
The Court asserts, however, that the decisions of this Court have placed a gloss on the phrase “within their respective jurisdictions” in §2241 which allows jurisdiction in these cases. That is not so. In fact, the only case in point holds just the opposite (and just what the statute plainly says). That case is Eisentrager, but to fully understand its implications for the present dispute, I must also discuss our decisions in the earlier case of Ahrens v. Clark, 335 U. S. 188 (1948), and the later case of Braden.
In Ahrens, the Court considered “whether the presence within the territorial jurisdiction of the District Court of the person detained is prerequisite to filing a petition for a writ of habeas corpus.” 335 U. S., at 189 (construing 28 U. S. C. §452, the statutory precursor to §2241). The Ahrens detainees were held at Ellis Island, New York, but brought their petitions in the District Court for the District of Columbia. Interpreting “within their respective jurisdictions,” the Court held that a district court has jurisdiction to issue the writ only on behalf of petitioners detained within its territorial jurisdiction. It was “not sufficient . . . that the jailer.or custodian alone be found in the jurisdiction.” 335 U. S., at 190.
Ahrens explicitly reserved “the question of what process, if any, a person confined in an area not subject to the jurisdiction of any district court may employ to assert federal rights.” Id., at 192, n. 4. That question, the same question *491presented to this Court today, was shortly thereafter resolved in Eisentrager insofar as noncitizens are concerned. Eisentrager involved petitions for writs of habeas corpus filed in the District Court for the District of Columbia by German nationals imprisoned in Landsberg Prison, Germany. The District Court, relying on Ahrens, dismissed the petitions because the petitioners were not located within its territorial jurisdiction. The Court of Appeals reversed. According to the Court today, the Court of Appeals “implicitly conceded that the District Court lacked jurisdiction under the habeas statute as it had been interpreted in Ahrens,” and “[i]n essence . . . concluded that the habeas statute, as construed in Ahrens, had created an unconstitutional gap that had to be filled by reference to ‘fundamentals.’” Ante, at 477, 478. That is not so. The Court of Appeals concluded that there was statutory jurisdiction. It arrived at that conclusion by applying the canon of constitutional avoidance: “[I]f the existing jurisdictional act be construed to deny the writ to a person entitled to it as a substantive right, the act would be unconstitutional. It should be construed, if possible, to avoid that result.” Eisentrager v. Forrestal, 174 F. 2d 961, 966 (CADC 1949). In cases where there was no territorial jurisdiction over the detainee, the Court of Appeals held, the writ would lie at the place of a respondent with directive power over the detainee. “It is not too violent an interpretation of ‘custody’ to construe it as including those who have directive custody, as well as those who have immediate custody, where such interpretation is necessary to comply with constitutional requirements.. . . The statute must be so construed, lest it be invalid as constituting a suspension of the writ in violation of the constitutional provision.” Id., at 967 (emphasis added).2
*492This Court’s judgment in Eisentrager reversed the Court of Appeals. The opinion was largely devoted to rejecting the lower court’s constitutional analysis, since the doctrine of constitutional avoidance underlay its statutory.conclusion. But the opinion had to pass judgment on whether the statute granted jurisdiction, since that was the basis for the judgments of both lower courts. A conclusion of no constitutionally conferred right would obviously not support reversal of a judgment that rested upon a statutorily conferred right.3 *493And absence of a right to the writ under the clear wording of the habeas statute is what the Eisentrager opinion held: “Nothing in the text of the Constitution extends such a right, nor does anything in our statutes” 339 U. S., at 768 (emphasis added). “[T]hese prisoners at no relevant time were within any territory over which the United States is sovereign, and the scenes of their offense, their capture, their trial and their punishment were all beyond the territorial jurisdiction of any court of the United States” Id., at 777-778. See also id., at 781 (concluding that “no right to the writ of habeas corpus appears”); id., at 790 (finding “no basis for invoking federal judicial power in any district”). The brevity of the Court’s statutory analysis signifies nothing more than that the Court considered it obvious (as indeed it is) that, unaided by the canon of constitutional avoidance, the statute did not confer jurisdiction over an alien detained outside the territorial jurisdiction of the courts of the United States.
Eisentrager]s directly-on-point statutory holding makes it exceedingly difficult for the Court to reach the result it desires today. To do so neatly and cleanly, it must either argue that our decision in Braden overruled Eisentrager, or admit that it is overruling Eisentrager. The former course would not pass the laugh test, inasmuch as Braden dealt with a detainee held within the territorial jurisdiction of a district court, and never mentioned Eisentrager. And the latter course would require the Court to explain why our almost categorical rule of stare decisis in statutory cases should be set aside in order to complicate the present war, and, having set it aside, to explain why the habeas statute does not mean what it plainly says. So instead the Court tries an oblique course: “Braden,” it claims, “overruled the statutory predi*494cate to Eisentrager’s holding,” ante, at 479 (emphasis added), by which it means the statutory analysis of Ahrens. Even assuming, for the moment, that Braden overruled some aspect of Ahrens, inasmuch as Ahrens did not pass upon any of the statutory issues decided by Eisentrager, it is hard to see how any of that case’s “statutory predicate” could have been impaired.
But in fact Braden did not overrule Ahrens; it distinguished Ahrens. Braden dealt with a habeas petitioner incarcerated in Alabama. The petitioner filed an application for a writ of habeas corpus in Kentucky, challenging an indictment that had been filed against him in that Commonwealth and naming as respondent the Kentucky court in which the proceedings were pending. This Court held that Braden was in custody because a detainer had been issued against him by Kentucky, and was being executed by Alabama, serving as an agent for Kentucky. We found that jurisdiction existed in Kentucky for Braden’s petition challenging the Kentucky detainer, notwithstanding his physical confinement in Alabama. Braden was careful to distinguish that situation from the general rule established in Ahrens.
“A further, critical development since our decision in Ahrens is the emergence of new classes of prisoners who are able to petition for habeas corpus because of the adoption of a more expansive definition of the ‘custody’ requirement of the habeas statute. The overruling of McNally v. Hill, 293 U. S. 131 (1934), made it possible for prisoners in custody under one sentence to attack a sentence which they had not yet begun to serve. And it also enabled a petitioner held in one State to attack a detainer lodged against him by another State. In such a case, the State holding the prisoner in immediate confinement acts as agent for the demanding State, and the custodian State is presumably indifferent to the resolution of the prisoner’s attack on the detainer. Here, for example, the petitioner is confined in Alabama, but his *495dispute is with the Commonwealth of Kentucky, not the State of Alabama. Under these circumstances, it would serve no useful purpose to apply the Ahrens rule and require that the action be brought in Alabama.” 410 U. S., at 498-499 (citations and footnotes omitted; emphases added).
This cannot conceivably be construed as an overturning of the Ahrens rule in other circumstances. See also Braden, supra, at 499-500 (noting that Ahrens does not establish “an inflexible jurisdictional rule dictating the choice of an inconvenient forum even in a class of cases which could not have been foreseen at the time of that decision” (emphasis added)). Thus, Braden stands for the proposition, and only the proposition, that where a petitioner is in custody in multiple jurisdictions within the United States, he may seek a writ of ha-beas corpus in a jurisdiction in which he suffers legal confinement, though not physical confinement, if his challenge is to that legal confinement. Outside that class of cases, Braden did not question the general rule of Ahrens (much less that of Eisentrager). Where, as here, present physical custody is at issue, Braden is inapposite, and Eisen-trager unquestionably controls.4
*496The considerations of forum convenience that drove the analysis in Braden do not call into question Eisentrager’s holding. The Braden opinion is littered with venue reasoning of the following sort: “The expense and risk of transporting the petitioner to the Western District of Kentucky, should his presence at a hearing prove necessary, would in all likelihood be outweighed by the difficulties of transporting records and witnesses from Kentucky to the district where petitioner is confined.” 410 U. S., at 494. Of course nothing could be more inconvenient than what the Court (on the alleged authority of Braden) prescribes today: a domestic hearing for persons held abroad, dealing with events that transpired abroad.
Attempting to paint Braden as a refutation of Ahrens (and thereby, it is suggested, Eisentrager), today’s Court imprecisely describes Braden as citing with approval post-Ahrens cases in which “habeas petitioners” located overseas were allowed to proceed (without consideration of the jurisdictional issue) in the District Court for the District of Columbia. Ante, at 479. In fact, what Braden said is that “[wjhere American citizens confined overseas (and thus outside the territory of any district court) have sought relief in habeas corpus, we have held, if only implicitly, that the *497petitioners’ absence from the district does not present a jurisdictional obstacle to the consideration of the claim.” 410 U. S., at 498 (emphasis added). Of course “the existence of unaddressed jurisdictional defects has no precedential, effect,” Lewis v. Casey, 518 U. S. 843, 352, n. 2 (1996) (citing cases), but we need not “overrule” those implicit holdings to decide these cases. Since Eisentrager itself made an exception for such cases, they in no way impugn its holding. “With the citizen,” Eisentrager said, “we are now little concerned, except to set his case apart as untouched by this decision and to take measure of the difference between his status and that of all categories of aliens.” 339 U. S., at 769. The constitutional doubt that the Court of Appeals in Eisen-trager had erroneously attributed to the lack of habeas for an alien abroad might indeed exist with regard to a citizen abroad — justifying a strained construction of the habeas statute, or (more honestly) a determination of constitutional right to habeas. Neither party to the present case challenges the atextual extension of the habeas statute to United States citizens held beyond the territorial jurisdictions of the United States courts; but the possibility of one atextual exception thought to be required by the Constitution is no justification for abandoning the clear application of the text to a situation in which it raises no constitutional doubt.
The reality is this: Today’s opinion, and today’s opinion alone, overrules Eisentrager; today’s opinion, and today’s opinion alone, extends the habeas statute, for the first time, to aliens held beyond the sovereign territory of the United States and beyond the territorial jurisdiction of its courts. No reasons are given for this result; no acknowledgment of its consequences made. By spurious reliance on Braden the Court evades explaining why stare decisis can be disregarded, and why Eisentrager was wrong. Normally, we consider the interests of those who have relied on our decisions. Today, the Court springs a trap on the Executive, subjecting Guantanamo Bay to the oversight of the federal *498courts even though it has never before been thought to be within their jurisdiction — and thus making it a foolish place to have housed alien wartime detainees.
II
In abandoning the venerable statutory line drawn in Ei-sentrager, the Court boldly extends the scope of the habeas statute to the four corners of the earth. Part III of its opinion asserts that Braden stands for the proposition that “a district court acts ‘within [its] respective jurisdiction’ within the meaning of §2241 as long as ‘the custodian can be reached by service of process.’” Ante, at 478-479. Endorsement of that proposition is repeated in Part IV. Ante, at 483-484 (“Section 2241, by its terms, requires nothing more [than the District Court’s jurisdiction over petitioners’ custodians]”).
The consequence of this holding, as applied to aliens outside the country, is breathtaking. It permits an alien captured in a foreign theater of active combat to bring a § 2241 petition against the Secretary of Defense. Over the course of the last century, the United States has held millions of alien prisoners abroad. See, e. g., Department of Army, G. Lewis & J. Mewha, History of Prisoner of War Utilization by the United States Army 1776-1945, Pamphlet No. 20-213, p. 244 (1955) (noting that, “[b]y the end of hostilities [in World War II], U. S. forces had in custody approximately two million enemy soldiers”). A great many of these prisoners would no doubt have complained about the circumstances of their capture and the terms of their confinement. The military is currently detaining over 600 prisoners at Guantanamo Bay alone; each detainee undoubtedly has complaints — real or contrived — about those terms and circumstances. The Court’s unheralded expansion of federal-court jurisdiction is not even mitigated by a comforting assurance that the legion of ensuing claims will be easily resolved on the merits. To the contrary, the Court says that the “[petitioners’ allegations ... unquestionably describe ‘custody in violation *499of the Constitution or laws or treaties of the United States.’ ” Ante, at 483, n. 15 (citing United States v. Verdugo-Urquidez, 494 U. S. 259, 277-278 (1990) (Kennedy, J., concurring)). From this point forward, federal courts will entertain petitions from these prisoners, and others like them around the world, challenging actions and events far away, and forcing the courts to oversee one aspect of the Executive’s conduct of a foreign war.
Today’s carefree Court disregards, without a word of acknowledgment, the dire warning of a more circumspect Court in Eisentrager:
“To grant the writ to these prisoners might mean that our army must transport them across the seas for hearing. This would require allocation for shipping space, guarding personnel, billeting and rations. It might also require transportation for whatever witnesses the prisoners desired to call as well as transportation for those necessary to defend legality of the sentence.' The writ, since it is held to be a matter of right, would be equally available to enemies during active hostilities as in the present twilight between war and peace. Such trials would hamper the war effort and bring aid and comfort to the enemy. They would diminish the prestige of our commanders, not only with enemies but with wavering neutrals. It would be difficult to devise more effective fettering of a field commander than to allow the very enemies he is ordered to reduce to submission to call him to account in his own civil courts and divert his efforts and attention from the military offensive abroad to the legal defensive at home. Nor is it unlikely that the result of such enemy litigiousness would be a conflict between judicial and military opinion highly comforting to enemies of the United States.” 339 U. S., at 778-779.
These results should not be brought about lightly, and certainly not without a textual basis in the statute and on the *500strength of nothing more than a decision dealing with an Alabama prisoner’s ability to seek habeas in Kentucky.
Ill
Part IV of the Court’s opinion, dealing with the status of Guantanamo Bay, is a puzzlement. The Court might have made an effort (a vain one, as I shall discuss) to distinguish Eisentrager on the basis of a difference between the status of Landsberg Prison in Germany and Guantanamo Bay Naval Base. But Part III flatly rejected such an approach, holding that the place of detention of an alien has no bearing on the statutory availability of habeas relief, but “is strictly relevant only to the question of the appropriate forum.” Ante, at 479. That rejection is repeated at the end of Part IV: “In the end, the answer to the question presented is clear.... No party questions the District Court’s jurisdiction over petitioners’ custodians. .. . Section 2241, by its terms, requires nothing more.” Ante, at 483-484. Once that has been said, the status of Guantanamo Bay is entirely irrelevant to the issue here. The habeas statute is (according to the Court) being applied domestically, to “petitioners’ custodians,” and the doctrine that statutes are presumed to have no extraterritorial effect simply has no application.
Nevertheless, the Court spends most of Part IV rejecting respondents’ invocation of that doctrine on the peculiar ground that it has no application to Guantanamo Bay. Of course if the Court is right about that, not only §2241 but presumably all United States law applies there — including, for example, the federal cause of action recognized in Bivens v. Six Unknown Fed. Narcotics Agents, 403 U. S. 388 (1971), which would allow prisoners to sue their captors for damages. Fortunately, however, the Court’s irrelevant discussion also happens to be wrong.
The Court gives only two reasons why the presumption against extraterritorial effect does not apply to Guantanamo Bay. First, the Court says (without any further elaboration) *501that “the United States exercises ‘complete jurisdiction and control’ over the Guantanamo Bay Naval Base [under the terms of a 1903 lease agreement], and may continue to exercise such control permanently if it so chooses [under the terms of a 1934 Treaty].” Ante, at 480; see ante, at 471. But that lease agreement explicitly recognized “the continuance of the ultimate sovereignty of the Republic of Cuba over the [leased areas],” Lease of Lands for Coaling and Naval Stations, Feb. 23, 1903, U. S.-Cuba, Art. III, T. S. No. 418, and the Executive Branch — whose head is “exclusively responsible” for the “conduct of diplomatic and foreign affairs,” Eisentrager, supra, at 789 — affirms that the lease and treaty do not render Guantanamo Bay the sovereign territory of the United States, see Brief for Respondents 21.
The Court does not explain how “complete jurisdiction and control” without sovereignty causes an enclave to be part of the United States for purposes of its domestic laws. Since “jurisdiction and control” obtained through a lease is no different in effect from “jurisdiction and control” acquired by lawful force of arms, parts of Afghanistan and Iraq should logically be regarded as subject to our domestic laws. Indeed, if “jurisdiction and control” rather than sovereignty were the test, so should the Landsberg Prison in Germany, where the United States held the Eisentrager detainees.
The second and last reason the Court gives for the proposition that domestic law applies to Guantanamo Bay is the Solicitor General’s concession that there would be habeas jurisdiction over a United States citizen in Guantanamo Bay. “Considering that the statute draws no distinction between Americans and aliens held in federal custody, there is little reason to think that Congress intended the geographical coverage of the statute to vary depending on the detainee’s citizenship.” Ante, at 481. But the reason the Solicitor General conceded there would be jurisdiction over a detainee who was a United States citizen had nothing to do with the special status of Guantanamo Bay: “Our answer to that ques*502tion, Justice Souter, is that citizens of the United States, because of their constitutional circumstances, may have greater rights with respect to the scope and reach of the Habeas Statute as the Court has or would interpret it.” Tr. of Oral Arg. 40. See also id., at 27-28. And that position — the position that United States citizens throughout the world may be entitled to habeas corpus rights — is precisely the position that this Court adopted in Eisentrager, see 339 U. S., at 769-770, even while holding that aliens abroad did not have habeas corpus rights. Quite obviously, the Court’s second reason has no force whatever.
The last part of the Court’s Part IV analysis digresses from the point that the presumption against extraterritorial application does not apply to Guantanamo Bay. Rather, it is directed to the, contention that the Court’s approach to habeas jurisdiction — applying it to aliens abroad — is “consistent with the historical reach of the writ.” Ante, at 481. None of the authorities it cites comes close to supporting that claim. Its first set of authorities involves claims by aliens detained in what is indisputably domestic territory. Ante, at 481-482, n. 11. Those cases are irrelevant because they do not purport to address the territorial reach of the writ. The remaining eases involve issuance of the writ to “'exempt jurisdictions’” and “other dominions under the sovereign’s, control.” Ante, at 482, and nn. 12-13. These cases are in-apposite for two reasons: Guantanamo Bay is not a sovereign dominion, and even if it were, jurisdiction would be limited to subjects.
“Exempt jurisdictions” — the Cinque Ports and Counties Palatine (located in modern-day England) — were local franchises granted by the Crown. See 1 W. Holdsworth, History of English Law 108, 532 (7th ed. rev. 1956); 3 W. Blackstone, Commentaries on the Laws of England 78-79 (1768) (hereinafter Blackstone). These jurisdictions were “exempt” in the sense that the Crown had ceded management of municipal affairs to local authorities, whose courts had exclusive juris*503diction over private disputes among residents (although review was still available in the royal courts by writ of error). See id., at 79. Habeas jurisdiction nevertheless extended to those regions on the theory that the delegation of the King’s authority did not include his own prerogative writs. Ibid.; R. Sharpe, Law of Habeas Corpus 188-189 (2d ed. 1989) (hereinafter Sharpe). Guantanamo Bay involves no comparable local delegation of pre-existing sovereign authority.
The cases involving “other dominions under the sovereign’s control” fare no better. These cases stand only for the proposition that the writ extended to dominions of the Crown outside England proper. The authorities relating to Jersey and the other Channel Islands, for example, see ante, at 482, n. 13, involve territories that are “dominions of the crown of Great Britain” even though not “part of the kingdom of England,” 1 Blackstone 102-105 (1765), much as were the colonies in America, id., at 104-105, and Scotland, Ireland, and Wales, id., at 93. See also King v. Cowle, 2 Burr. 834, 853-854, 97 Eng. Rep. 587, 598 (K. B. 1759) (even if Ber-wick was “no part of the realm of England,” it was still a “dominion of the Crown”). All of the dominions in the cases the Court cites — and all of the territories Blackstone lists as dominions, see 1 Blackstone 93-106 — are the sovereign territory of the Crown: colonies, acquisitions and conquests, and so on. It is an enormous extension of the term to apply it to installations merely leased for a particular use from another nation that still retains ultimate sovereignty.
The Court’s historical analysis fails for yet another reason: To the extent the writ’s “extraordinary territorial ambit” did extend to exempt jurisdictions, outlying dominions, and the like, that extension applied only to British subjects. The very sources the majority relies on say so: Sharpe explains the “broader ambit” of the writ on the ground that it is “said to depend not on the ordinary jurisdiction of the court for its effectiveness, but upon the authority of the sovereign over *504all her subjects” Sharpe 188 (emphasis added). Likewise, Blackstone explained that the writ “run[s] into all parts of the king’s dominions” because “the king is at all times entitled to have an account why the liberty of any of his subjects is restrained.” 3 Blackstone 131 (emphasis added). Ex parte Mwenya, [1960] 1 Q. B. 241 (C. A.), which can hardly be viewed as evidence of the historic scope of the writ, only confirms the ongoing relevance of the sovereign-subject relationship to the scope of the writ. There, the question was whether “the Court of Queen’s Bench [can] be debarred from making an order in favour of a British citizen unlawfully or arbitrarily detained” in Northern Rhodesia, which was at the time a protectorate of the Crown. Id., at 300 (Lord Evershed, M. R.). Each judge made clear that the detainee’s status as a subject was material to the resolution of the case. See id., at 300, 302 (Lord Evershed, M. R.); id., at 305 (Romer, L. J.) (“[I]t is difficult to see why the sovereign should be deprived of her right to be informed through her High Court as to the validity of the detention of her subjects.in that territory”); id., at 311 (Sellers, L. J.) (“I am not prepared to say, as we are solely asked to say on this appeal, that the English courts have no jurisdiction in any circumstances to entertain an application for a writ of habeas corpus ad subjiciendum in respect of an unlawful detention of a British subject in a British protectorate”). None of the exempt-jurisdiction or dominion cases the Court cites involves someone not a subject of the Crown.
The rule against issuing the writ to aliens in foreign lands was still the law when, in In re Ning Yi-Ching, 56 T. L. R. 3 (K. B. Vac. Ct. 1939), an English court considered the habeas claims of four Chinese subjects detained on criminal charges in Tientsin, China, an area over which Britain had by treaty acquired a lease and “therewith exercised certain rights of administration and control.” Id., at 4. The court held that Tientsin was a foreign territory, and that the writ would not *505issue to a foreigner detained there. The Solicitor-General had argued that “[t]here was no case on record in which a writ of habeas corpus had been obtained on behalf of a foreign subject on foreign territory,” id., at 5, and the court “listened in vain for a case in which the writ of habeas corpus had issued in respect of a foreigner detained in a part of the world which was not a part of the King’s dominions or realm,” id., at 6.5
In sum, the Court’s treatment of Guantanamo Bay, like its treatment of §2241, is a wrenching departure from precedent.6
*506* * *
Departure from our rule of stare decisis in statutory cases is always extraordinary; it ought to be unthinkable when the departure has a potentially harmful effect upon the Nation’s conduct of a war. The Commander in Chief and his subordinates had every reason to expect that.the internment of combatants at Guantanamo Bay would not have the consequence of bringing the cumbersome machinery of our domestic courts into military affairs. Congress is in session. If it wished to change federal judges’ habeas jurisdiction from what this Court had previously held that to be, it could have done so. And it could have done so by intelligent revision of the statute,7 instead of by today’s clumsy, countertex-tual reinterpretation that confers upon wartime prisoners greater habeas rights than domestic detainees. The latter must challenge their present physical confinement in the district of their confinement, see Rumsfeld v. Padilla, ante, p. 426, whereas under today’s strange holding Guantanamo Bay detainees can petition in any of the 94 federal judicial districts. The fact that extraterritorially located detainees lack the district of detention that the statute requires has been converted from a factor that precludes their ability to bring a petition at all into a factor that frees them to petition wherever they wish — and, as a result, to forum-shop. For this Court to create such a monstrous scheme in time of war, and in frustration of our military commanders’ reliance upon clearly stated prior law, is judicial adventurism of the worst sort. I dissent.
See Tr. of Oral Arg. 5 (“Question: And you don’t raise the issue of any potential jurisdiction on the basis of the Constitution alone. We are here debating the jurisdiction under the Habeas Statute, is that right? [Answer]: That’s correct...”).
The parties’ submissions to the Court in Johnson v. Eisentrager, 339 U. S. 763 (1950), construed the Court of Appeals’ decision as I do. See Pet. for Cert., O. T. 1949, No. 306, pp. 8-9 (“[T]he court felt constrained to construe the habeas corpus jurisdictional statute — despite its reference to *492the ‘respective jurisdictions’ of the various courts and the gloss put on that terminology in the Ahrens and previous decisions — to permit a petition to be filed in the district court with territorial jurisdiction over the officials who have directive authority over the immediate jailer in Germany”); Brief for Respondent, O. T. 1949, No. 306, p. 9 (“Respondent contends that the U. S. Court of Appeals . .. was correct in its holding that the statute, 28 U. S. C. 2241, provides that the U. S. District Court for the District of Columbia has jurisdiction to entertain the petition for a writ of habeas corpus in the case at bar”). Indeed, the briefing in Eisentrager was mainly devoted to the question whether there was statutory jurisdiction. See, e. g., Brief for Petitioner, O. T. 1949, No. 306, pp. 15-59; Brief for Respondent, O. T. 1949, No. 306, at 9-27, 38-49.
The Court does not seriously dispute my analysis of the Court of Appeals’ holding in Eisentrager. Instead, it argues that this Court in Eisentrager “understood the Court of Appeals’ decision to rest on constitutional and not statutory grounds. ” Ante, at 478, n. 8. That is inherently implausible, given that the Court of Appeals’ opinion clearly reached a statutory holding, and that both parties argued the case to this Court on that basis, see n. 2, supra. The only evidence of misunderstanding the Court adduces today is the Eisentrager Court’s description of the Court of Appeals’ reasoning as “that, although no statutory jurisdiction of such cases is given, courts must be held to possess it as part of the judicial power of the United States ....” 339 U. S., at 767. That is no misunderstanding, but an entirely accurate description of the Court of Appeals’ reasoning— the penultimate step of that reasoning rather than its conclusion. The Court of Appeals went on to hold that, in light of the constitutional imperative, the statute should be interpreted as supplying jurisdiction. See Eisentrager v. Forrestal, 174 F. 2d 961, 965-967 (CADC 1949). This Court in Eisentrager undoubtedly understood that, which is why it immediately followed the foregoing description with a description of the Court of Appeals’ conclusion tied to the language of the habeas statute: *493“[W]here deprivation of liberty by an official act occurs outside the territorial jurisdiction of any District Court, the petition will lie in the District Court which has territorial jurisdiction over officials who have directive power over the immediate jailer.” 339 U. S., at 767.
The Court points to Court of Appeals cases that have described Braden as “overruling” Ahrens. See ante, at 479-480, n. 9. Even if that description (rather than what I think the correct one, “distinguishing”) is accepted, it would not support the Court’s view that Ahrens was overruled with regard to the point on which Eisentrager relied. The ratio decidendi of Braden does not call into question the principle of Ahrens applied in Eisentrager: that habeas challenge to present physical confinement must be made in the district where the physical confinement exists. The Court is unable to produce a single authority that agrees with its conclusion that Braden overruled Eisentrager.
Justice Kennedy recognizes that Eisentrager controls, ante, at 485 (opinion concurring in judgment), but misconstrues that opinion. He thinks it makes jurisdiction under the habeas statute turn on the circumstances of the detainees’ confinement — including, apparently, the availability of legal proceedings and the length of detention, see ante, at 487-*496488. The Eisentrager Court mentioned those circumstances, however, only in the course of its constitutional analysis, and not in its application of the statute. It is quite impossible to read §2241 as conditioning its geographic scope upon them. Among the consequences of making jurisdiction turn upon circumstances of confinement are (1) that courts would always have authority to inquire into circumstances of confinement, and (2) that the Executive would be unable to know with certainty that any given prisoner-of-war camp is immune from writs of habeas corpus. And among the questions this approach raises: When does definite detention become indefinite? How much process will suffice to stave off jurisdiction? If there is a terrorist attack at Guantanamo Bay, will the area suddenly fall outside the habeas statute because it is no longer “far removed from any hostilities,” ante, at 487? Justice Kennedy’s approach provides enticing law-school-exam imponderables in an area where certainty is called for.
The Court argues at some length that Ex parte Mwenya, [1960] 1 Q. B. 241 (C. A.), calls into question my reliance on In re Ning Yi-Ching. See ante, at 15, n. 14. But as I have explained, see supra, at 504, Mwenya dealt with a British subject and the court went out of its way to explain that its expansive description of the scope of the writ was premised on that fact. The Court cites not a single case holding that aliens held outside the territory of the sovereign were within reach of the writ.
The Court grasps at two other bases for jurisdiction: the Alien Tort Statute (ATS), 28 U. S. C. § 1350, and the federal-question statute, § 1331. The former is not presented to us. The ATS, while invoked below, was repudiated as a basis for jurisdiction by all petitioners, either in their petition for certiorari, in their briefing before this Court, or at oral argument. See Pet. for Cert. in No. 03-334, p. 2, n. 1 (“Petitioners withdraw any reliance on the Alien Tort Claims Act. . .”); Brief for Petitioners in No. 03-343, p. 13; Tr. of Oral Arg. 6.
With respect to § 1331, petitioners assert a variety of claims arising under the Constitution, treaties, and laws of the United States. In Eisen-trager, though the Court’s holding focused on §2241, its analysis spoke more broadly: “We have pointed out that the privilege of litigation has been extended to aliens, whether friendly or enemy, only because permitting their presence in the country implied protection. No such basis can be invoked here, for these prisoners at no relevant time were within any territory over which the United States is sovereign, and the scenes of their offense, their capture, their trial and their punishment were all beyond the territorial jurisdiction of any court of the United States.” 339 U. S., at 777-778. That reasoning dooms petitioners’ claims under § 1331, at least where Congress has erected a jurisdictional bar to their raising such claims in habeas.
It could, for example, provide for jurisdiction by placing Guantanamo Bay within the territory of an existing district court; or by creating a district court for Guantanamo Bay, as it did for the Panama Canal Zone, see 22 U. S. C. § 8841(a) (repealed 1979).