After the Chicago River flooded a freight tunnel under the river and the basements of numerous buildings, petitioner corporation and other victims brought tort actions in state court against respondent Great Lakes Dredge & Dock Co. and petitioner Chicago. They claimed that in the course of driving piles from a barge into the river bed months earlier, Great Lakes had negligently weakened the tunnel, which had been improperly maintained by the city. Great Lakes then filed this action, invoking federal admiralty jurisdiction and seeking, inter alia, the protection of the Limitation of Vessel Owner's Liability Act. That Act would permit the admiralty court to decide whether Great Lakes had committed a tort and, if so, to limit its liability to the value of the barges and tug involved if the tort was committed without the privity or knowledge of the vessels' owner. The District Court dismissed the suit for lack of admiralty jurisdiction, but the Court of Appeals reversed.

Held:

The District Court has federal admiralty jurisdiction over Great Lakes's Limitation Act suit. Pp. 3-21.

(a) A party seeking to invoke such jurisdiction over a tort claim must satisfy conditions of both location and connection with maritime activity. In applying the location test, a court must determine whether the tort occurred on navigable water or whether injury suffered on land was caused by a vessel on navigable water. 46 U.S.C. App. 740. In applying the connection test, a court first Page II must assess the "general features of the type of incident involved" to determine if the incident has "a potentially disruptive impact on maritime commerce." Sisson v. Ruby, 497 U.S. 358, 363 , 364, n. 2. If so, the court must determine whether the character of the activity giving rise to the incident shows a substantial relationship to traditional maritime activity. Id., at 365, 364, and n. 2. Pp. 3-6.

(b) The location test is readily satisfied here. The alleged tort was committed on a navigable river, and petitioners do not seriously dispute that Great Lakes's barge is a "vessel" for admiralty tort purposes. There is no need or justification for imposing an additional jurisdictional requirement that the damage done must be close in time and space to the activity that caused it. A nonremoteness requirement is not supported by the Extension of Admiralty Jurisdiction Act's language, and the phrase "caused by" used in that Act indicates that the proper standard is proximate cause. Gutierrez v. Waterman S. S. Corp., 373 U.S. 206, 210 , distinguished. Pp. 6-10.

(c) The maritime connection test is also satisfied here. The incident's "general features" may be described as damage by a vessel in navigable water to an underwater structure. There is little question that this is the kind of incident that has "a potentially disruptive impact on maritime commerce." Damaging the structure could lead to a disruption in the water course itself and, as actually happened here, could lead to restrictions on navigational use during repairs. There is also no question that the activity giving rise to the incident - repair or maintenance work on a navigable waterway performed from a vessel - shows a substantial relationship to traditional maritime activity. Even the assertion that the city's alleged failure to properly maintain and operate the tunnel system was a proximate cause of the flood damage does not take this case out of admiralty. Under Sisson, the substantial relationship test is satisfied when at least one alleged tortfeasor was engaging in activity substantially related to traditional maritime activity and such activity is claimed to have been a proximate cause of the incident. There is no merit to the argument that the activity should be characterized at a hypergeneralized level, such as "repair and maintenance," to eliminate any hint of maritime connection, or to the argument that Sisson is being given too expansive a reading. Pp. 10-16.

(d) There are theoretical, as well as practical, reasons to reject the city's proposed multifactor test for admiralty jurisdiction where most of the victims, and one of the tortfeasors, are land based. The Sisson tests are directed at the same objectives invoked to support a multifactor test, the elimination of admiralty jurisdiction where the rationale for the jurisdiction does not support it. In the Extension Page III Act, Congress has already made a judgment that a land-based victim may properly be subject to admiralty jurisdiction; surely a land-based joint tortfeasor has no claim to supposedly more favorable treatment. Moreover, contrary to the city's position, exercise of admiralty jurisdiction does not result in automatic displacement of state law. A multifactor test would also be hard to apply, jettisoning relative predictability for the open-ended rough-and-tumble of factors, inviting complex argument in a trial court and a virtually inevitable appeal. Pp. 16-21.

SOUTER, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and O'CONNOR, KENNEDY, and GINSBURG, JJ., joined. O'CONNOR, J., filed a concurring opinion. THOMAS, J., filed an opinion concurring in the judgment, in which SCALIA, J., joined. STEVEN and BREYER, JJ., took no part in the decision of the case.

JUSTICE SOUTER delivered the opinion of the Court.

On April 13, 1992, water from the Chicago River poured into a freight tunnel running under the river and thence into the basements of buildings in the downtown Chicago Loop. Allegedly, the flooding resulted from events several months earlier, when the respondent Great Lakes Dredge and Dock Company had used a crane, sitting on a barge in the river next to a bridge, to drive piles into the river bed above the tunnel. The issue before us is whether a court of the United States has admiralty jurisdiction to determine and limit the extent of Great Lakes's tort liability. We hold the case to be within federal admiralty jurisdiction.

In August and September 1991, Great Lakes replaced the pilings around the piers projecting into the river and supporting the Kinzie Street Bridge. After towing the crane-carrying barge into position near one of the piers, Great Lakes's employees secured the barge to the river bed with spuds, or long metal legs that project down from the barge and anchor it. The workers then used the crane on the barge to pull up old pilings, stow them on the other barge, and drive new pilings into the river bed around the piers. About seven months later, an eddy formed in the river near the bridge as the collapsing walls or ceiling of a freight tunnel running under the river opened the tunnel to river water, which flowed through to flood buildings in the Loop.

After the flood, many of the victims brought actions in state court against Great Lakes and the city of Chicago, claiming that in the course of replacing the pilings Great Lakes had negligently weakened the tunnel structure, which Chicago (its owner) had not properly maintained. Great Lakes then brought this lawsuit in the United States District Court, invoking federal admiralty jurisdiction. Count I of the complaint seeks the protection of the Limitation of Vessel Owner's Liability Act (Limitation Act), 46 U.S.C. App. 181 et seq., a statute that would, in effect, permit the admiralty court to decide whether Great Lakes committed a tort and, if so, to limit Great Lakes's liability to the value of the vessels (the tug and two barges) involved if the tort was committed "without the privity or knowledge" of the vessels' owner, 46 U.S.C. App. 183(a). Counts II and III of Great Lakes's complaint ask for indemnity and contribution from the city for any resulting loss to Great Lakes.

The city, joined by petitioner Jerome B. Grubart, Inc., one of the state-court plaintiffs, filed a motion to dismiss this suit for lack of admiralty jurisdiction. Fed. Rule Civ. Proc. 12(b)(1). The District Court granted the motion, the Seventh Circuit reversed, Great Lakes Dredge & Dock Co. v. Chicago, 3 F.3d 225 (1993), and we granted certiorari, 510 U.S. ___ (1994). We now affirm.

The parties do not dispute the Seventh Circuit's conclusion that jurisdiction as to Counts II and III (indemnity and contribution) hinges on jurisdiction over the Count I claim. See 3 F.3d, at 231, n. 9; see also 28 U.S.C. 1367 (1988 ed., Supp. V) (supplemental jurisdiction); Fed. Rules Civ. Proc. 14(a) and (c) (impleader of third parties). Thus, the issue is simply whether or not a federal admiralty court has jurisdiction over claims that Great Lakes's faulty replacement work caused the flood damage.

The traditional test for admiralty tort jurisdiction asked only whether the tort occurred on navigablewaters. If it did, admiralty jurisdiction followed; if it did not, admiralty jurisdiction did not exist. See, e.g., Thomas v. Lane, 23 F. Cas. 957, 960 (No. 13902) (CC Me. 1813) (Story, J., on Circuit). This ostensibly simple locality test was complicated by the rule that the injury had to be "wholly" sustained on navigable waters for the tort to be within admiralty. The Plymouth, 3 Wall. 20, 34 (1866) (no jurisdiction over tort action brought by the owner of warehouse destroyed in a fire that started on board a ship docked nearby). Thus, admiralty courts lacked jurisdiction over, say, a claim following a ship's collision with a pier insofar as it injured the pier, for admiralty law treated the pier as an extension of the land. Martin v. West, 222 U.S. 191, 197 (1911); Cleveland T. & V. R. Co. v. Cleveland S. S. Co., 208 U.S. 316, 319 (1908).

This latter rule was changed in 1948, however, when Congress enacted the Extension of Admiralty Jurisdiction Act, 62 Stat. 496. The Act provided that

"[t]he admiralty and maritime jurisdiction of the United States shall extend to and include all cases of damage or injury, to person or property, caused by a vessel on navigable water, notwithstanding that such damage or injury be done or consummated on land." 46 U.S.C. App. 740.

After this congressional modification to gather the odd case into admiralty, the jurisdictional rule was qualified again in three decisions of this Court aimed at keepinga different class of odd cases out. In the first case, Executive Jet, supra, tort claims arose out of the wreck of an airplane that collided with a flock of birds just after take-off on a domestic flight and fell into the navigable waters of Lake Erie. We held that admiralty lacked jurisdiction to consider the claims. We wrote that "a purely mechanical application of the locality test" was not always "sensible" or "consonant with the purposes of maritime law, id., at 261, as when (for example) the literal and universal application of the locality rule would require admiralty courts to adjudicate tort disputes between colliding swimmers, id., at 255. We held that "claims arising from airplane accidents are not cognizable in admiralty" despite the location of the harm, unless the "the wrong bear[s] a significant relationship to traditional maritime activity." Id., at 268.

The second decision, Foremost Ins. Co. v. Richardson, 457 U.S. 668 (1982), dealt with tort claims arising out of the collision of two pleasure boats in a navigable river estuary. We held that admiralty courts had jurisdiction, id., at 677, even though jurisdiction existed only if "the wrong" had "a significant connection with traditional maritime activity," id., at 674. We conceded that pleasure boats themselves had little to do with the maritime commerce lying at the heart of the admiralty court's basic work, id., at 674-675, but we nonetheless found the necessary relationship in

"[t]he potential disruptive impact [upon maritime commerce] of a collision between boats on navigable waters, when coupled with the traditional concern that admiralty law holds for navigation . . . ." Id., at 675.

After Sisson, then, a party seeking to invoke federal admiralty jurisdiction pursuant to 28 U.S.C. 1333(1) over a tort claim must satisfy conditions both of location and of connection with maritime activity. A court applying the location test must determine whether the tort occurred on navigable water or whether injury suffered on land was caused by a vessel on navigable water. 46 U.S.C. App. 740. The connection test raises two issues. A court, first, must "assess the general features of the type of incident involved," 497 U.S., at 363 , to determine whether the incident has "a potentially disruptive impact on maritime commerce," id., at 364, n. 2. Second, a court must determine whether "the general character" of the "activity giving rise to the incident" shows a "substantial relationship to traditional maritime activity." Id., at 365, 364, and n. 2. We now apply the tests to the facts of this case.

It must also have done it "by a vessel." Even though the barge was fastened to the river bottom and was in use as a work platform at the times in question, at other times it was used for transportation. See 3 F.3d, at 229. Petitioners do not here seriously dispute the conclusion of each court below that the Great Lakes barge is, for admiralty tort purposes, a "vessel." The fact that the pile-driving was done with a crane makes no difference under the location test, given the maritime law that ordinarily treats an "appurtenance" attached to a vessel in navigable waters as part of the vessel itself. See, e.g., Victory Carriers, Inc. v. Law, 404 U.S. 202, 210 -211 (1971); Gutierrez, 373 U.S., at 209 -210.¹

Because the injuries suffered by Grubart and the other flood victims were caused by a vessel on navigable water, the location enquiry would seem to be at an end, "notwithstanding that such damage or injury [have been] done or consummated on land." 46 U.S.C. App. 740. Both Grubart and Chicago nonetheless ask us to subject the Extension Act to limitations not apparent from its text. While they concede that the Act refers to "all cases of damage or injury," they argue that "all" must not mean literally every such case, no matter how great the distance between the vessel's tortious activity and the resulting harm. They contend that, to be within the Act, the damage must be close in time and space to the activity that caused it: that it must occur "reasonably contemporaneously" with the negligent conduct and no "farther from navigable waters than the reach of the vessel, its appurtenances and cargo." For authority, they point to this Court's statement in Gutierrez, supra, that jurisdiction is present when the "impact" of the tortious activity "is felt ashore at a time and place not remote from the wrongful act." Id., at 210.²

The demerits of this argument lie not only in its want of textual support for its nonremoteness rule, but in its disregard of a less stringent but familiar proximity condition tied to the language of the statute. The Act uses the phrase "caused by," which more than one Court of Appeals has read as requiring what tort law has traditionally called "proximate causation." See, e.g., Pryor v. American President Lines, 520 F.2d 974, 979 (CA4 1975), cert. denied, 423 U.S. 1055 (1976); Adams v. Harris County, 452 F.2d 994, 996-997 (CA5 1971),cert. denied, 406 U.S. 968 (1972). This classic tort notion normally eliminates the bizarre, cf. Palsgraf v. Long Island R. Co., 248 N.Y. 339, 162 N. E. 99 (1928), and its use should obviate not only the complication but even the need for further temporal or spatial limitations. Nor is reliance on familiar proximate causation inconsistent with Gutierrez, which used its nonremote language, not to announce a special test, but simply to distinguish its own facts (the victim having slipped on beans spilling from cargo containers being unloaded from a ship) from what the Court called "[v]arious far-fetched hypotheticals," such as injury to someone slipping on beans that continue to leak from the containers after they had been shipped from Puerto Rico to a warehouse in Denver. 373 U.S., at 210 . See also Victory Carriers, supra, at 210-211.

The city responds by saying that, as a practical matter, the use of proximate cause as a limiting jurisdictional principle would undesirably force an admiralty court to investigate the merits of the dispute at the outset of a case when it determined jurisdiction.³  The argument, of course, assumes that the truth of jurisdictional allegations must always be determined with finality at the threshold of litigation, but that assumption is erroneous. Normal practice permits a party to establish jurisdiction at the outset of a case by means of a nonfrivolous assertion of jurisdictional elements, see, e.g., Bray v. Alexandria Women's Health Clinic, 506 U.S. ___, ___ (1993) (slip op., at 21); Bell v. Hood, 327 U.S. 678, 682 -683 (1946), and any litigation of a contested subject-matter jurisdictional fact issue occurs in comparatively summary procedure before a judge alone (as district from litigation of the same fact issue as an element of the cause of action, if the claim survives the jurisdictional objection). See 2A J. Moore & J. Lucas, Moore's Federal Practice  12.07[2. - 1] (2d ed. 1994); 5A C. Wright & A. Miller, Federal Practice and Procedure 1350 (2d ed. 1990). There is no reason why this should not be just as true for proximate causation as it is for the maritime nature of the tortfeasor's activity giving rise to the incident. See Sisson, 497 U.S., at 365 . There is no need or justification, then, for imposing an additional nonremoteness hurdle in the name of jurisdiction.

The first Sisson test turns, then, on a description of the incident at an intermediate level of possible generality. To speak of the incident as "fire" would have been too general to differentiate cases; at the other extreme, to have described the fire as damaging nothing but pleasure boats and their tie-up facilities would have ignored, among other things, the capacity of pleasure boats to endanger commercial shipping that happened to be nearby. We rejected both extremes and instead asked whether the incident could be seen within a class of incidents that posed more than a fanciful risk to commercial shipping.

Following Sisson, the "general features" of the incident at issue here may be described as damage by a vessel in navigable water to an underwater structure. So characterized, there is little question that this is the kind of incident that has a "potentially disruptive impact on maritime commerce." As it actually turned out in this case, damaging a structure beneath the river bed could lead to a disruption in the water course itself, App. 33 (eddy formed above the leak); and, again as it actually happened, damaging a structure so situated could lead to restrictions on the navigational use of the waterway during required repairs. See Pet. for Cert. in No. 93-1094, p. 22a (District Court found that after the flood "[t]he river remained closed for over a month," "[r]iver traffic ceased, several commuter ferries were stranded, and many barges could not enter the river system . . . because the river level was lowered to aid repair efforts"). Cf. Pennzoil Producing Co. v. Offshore Express, Inc., 943 F.2d 1465 (CA5 1991) (admiralty suit when vessel struck and ruptured gas pipeline and gas exploded); Marathon Pipe Line Co. v. Drilling Rig Rowan/Odessa, 761 F.2d 229, 233 (CA5 1985) (admiralty jurisdiction when vessel struck pipeline, "a fixed structure on the seabed"); Orange Beach Water, Sewer, and Fire Protection Authority v. M/V Alva, 680 F.2d 1374 (CA11 1982) (admiralty suit when vessel struck underwater pipeline).

In the second Sisson enquiry, we look to whether the general character of the activity giving rise to the incident shows a substantial relationship to traditional maritime activity. We ask whether a tortfeasor's activity, commercial or noncommercial, on navigable waters is so closely related to activity traditionally subject to admiralty law that the reasons for applying special admiralty rules would apply in the case at hand. Navigation of boats in navigable waters clearly falls within the substantial relationship, Foremost, 457 U.S., at 675 ; storing them at a marina on navigable waters is close enough, Sisson, supra, at 367; whereas in flying an airplane over the water, Executive Jet, 409 U.S., at 270 -271, as in swimming, id., at 255-256, the relationship is too attenuated.

On like reasoning, the "activity giving rise to the incident" in this case, Sisson, supra, at 364, should be characterized as repair or maintenance work on a navigable waterway performed from a vessel. Described in this way, there is no question that the activity is substantially related to traditional maritime activity, for barges and similar vessels have traditionally been engaged in repair work similar to what Great Lakes contracted to perform here. See, e.g., Shea v. Rev-Lyn Contracting Co., 868 F.2d 515, 518 (CA1 1989) (bridge repair by crane-carrying barge); Nelson v. United States, 639 F.2d 469, 472 (CA9 1980) (Kennedy, J.) (repair of wave suppressor from a barge); In re New York Dock Co., 61 F.2d 777 (CA2 1932) (pile driving from crane-carrying barge in connection with the building of a dock); In re P. Sanford Ross, Inc., 196 F. 921, 923-924 (EDNY 1912) (pile driving from crane-carrying barge close to water's edge), rev'd on other grounds, 204 F. 248 (CA2 1913); cf. In re The V-14813, 65 F.2d 789, 790 (CA5 1933) ("[t]here are many cases holding that a dredge, or a barge with a pile driver, employed on navigable waters, is subject to maritime jurisdiction"); Lawrence v. Flatboat, 84 F. 200 (SD Ala. 1897) (pile driving from crane-carrying barge in connection with the erection of bulkheads), aff'd sub nom. Southern Log Cart & Supply Co. v. Lawrence, 86 F. 907 (CA5 1898).

The city argues, to the contrary, that a proper application of the activity prong of Sisson would consider the city's own alleged failure at properly maintaining and operating the tunnel system that runs under the river. City Brief 48-49. If this asserted proximate cause of the flood victims' injuries were considered, the city submits, its failure to resemble any traditional maritime activity would take this case out of admiralty.

The city misreads Sisson, however, which did not consider the activities of the washer/dryer manufacturer, who was possibly an additional tortfeasor, and whose activities were hardly maritime; the activities of Sisson, the boat owner, supplied the necessary substantial relationship to traditional maritime activity. Likewise, in Foremost, we said that "[b]ecause the `wrong' here involves the negligent operation of a vessel on navigable waters, we believe that it has a sufficient nexus to traditional maritime activity to sustain admiralty jurisdiction . . . ." 457 U.S., at 674 . By using the word "involves," we made it clear that we need to look only to whether one of the arguably proximate causes of the incident originated in the maritime activity of a tortfeasor: as long as one of the putative tortfeasors was engaged in traditional maritime activity the allegedly wrongful activity will "involve" such traditional maritime activity and will meet the second nexus prong. Thus, even if we were to identify the "activity giving rise to the incident" as including the acts of the city as well as Great Lakes, admiralty jurisdiction would nevertheless attach. That result would be true to Sisson's requirement of a "substantial relationship" between the "activity giving rise to the incident" and traditional maritime activity. Sisson did not require, as the city in effect asserts, that there be a complete identity between the two. The substantial relationship test is satisfied when at least one alleged tortfeasor was engaging in activity substantially related to traditional maritime activity and such activity is claimed to have been a proximate cause of the incident.

Petitioners also argue that we might get a different result simply by characterizing the "activity" in question at a different level of generality, perhaps as "repair and maintenance," or, as "pile driving near a bridge." The city is, of course, correct that a tortfeasor's activity can be described at a sufficiently high level of generality to eliminate any hint of maritime connection, and if that were properly done Sisson would bar assertion of admiralty jurisdiction. But to suggest that such hyper-generalization ought to be the rule would convert Sisson into a vehicle for eliminating admiralty jurisdiction. Although there is inevitably some play in the joints in selecting the right level of generality when applying the Sisson test, the inevitable imprecision is not an excuse for whimsy. The test turns on the comparison of traditional maritime activity to the arguably maritime character of the tortfeasor's activity in a given case; the comparison would merely be frustrated by eliminating the maritime aspect of the tortfeasor's activity from consideration.⁴

Grubart makes an additional claim that Sisson is being given too expansive a reading. If the activity at issue here is considered maritime-related, it argues, then virtually "every activity involving a vessel on navigable waters" would be "a traditional maritime activity sufficient to invoke maritime jurisdiction." Grubart Brief 6. But this is not fatal criticism. This Court has not proposed any radical alteration of the traditional criteria for invoking admiralty jurisdiction in tort cases, but has simply followed the lead of the lower federal courts in rejecting a location rule so rigid as to extend admiralty to a case involving an airplane, not a vessel, engaged in an activity far removed from anything traditionally maritime. See Executive Jet, 409 U.S., at 268 -274; see also Peytavin v. Government Employees Ins. Co., 453 F.2d 1121, 1127 (CA5 1972) (no jurisdiction over claim for personal injury by motorist who was rear-ended while waiting for a ferry on a floating pontoon serving as the ferry's landing); Chapman v. Grosse Pointe Farms, 385 F.2d 962 (CA6 1967) (no admiralty jurisdiction over claim of swimmer who injured himself when diving off pier into shallow but navigable water). In the cases after Executive Jet, the Court stressed the need for a maritime connection, but found one in the navigation or berthing of pleasure boats, despite the facts that the pleasure boat activity took place near shore, where States have a strong interest in applying their own tort law, or was not on all fours with the maritime shipping and commerce that has traditionally made up the business of most maritime courts. Sisson, 497 U.S., at 367 ; Foremost, 457 U.S., at 675 . Although we agree with petitioners that these cases do not say that every tort involving a vessel on navigable waters falls within the scope of admiralty jurisdiction no matter what, they do show that ordinarily that will be so.⁵

"the totality of the circumstances reflects a federal interest in protecting maritime commerce sufficiently weighty to justify shifting what would otherwise be state-court litigation into federal court under the federal law of admiralty." City Brief 32.

Although the arguments are not frivolous, they do not persuade. It is worth recalling that the Sisson tests are aimed at the same objectives invoked to support a new multifactor test, the elimination of admiralty jurisdiction where the rationale for the jurisdiction does not support it. If the tort produces no potential threat to maritime commerce or occurs during activity lacking a substantial relationship to traditional maritime activity, Sisson assumes that the objectives of admiralty jurisdiction probably do not require its exercise, even if the location test is satisfied. If, however, the Sisson tests are also satisfied, it is not apparent why the need for admiralty jurisdiction in aid of maritime commerce somehow becomes less acute merely because land-based parties happen to be involved. Certainly Congress did not think a land-based party necessarily diluted the need for admiralty jurisdiction or it would have kept its hands off the primitive location test.

Of course one could claim it to be odd that under Sisson a land-based party (or more than one) may be subject to admiralty jurisdiction, but it would appear no less odd under the city's test that a maritime tortfeasor in the most traditional mould might be subject to state common-law jurisdiction. Other things being equal, it is not evident why the first supposed anomaly is worse than the second. But other things are not even equal. As noted just above, Congress has already made the judgment, in the Extension Act, that a land-based victim may properly be subject to admiralty jurisdiction. Surely a land-based joint tortfeasor has no claim to supposedly more favorable treatment.

Nor are these the only objections to the city's position. Contrary to what the city suggests, City Brief 10, 14-15, 25-26, 30, exercise of federal admiralty jurisdiction does not result in automatic displacement of state law. It is true that, "[w]ith admiralty jurisdiction comes the application of substantive admiralty law." East River S. S. Corp. v. Transamerica DeLaval Inc., 476 U.S. 858, 864 (1986). But, to characterize that law, as the city apparently does, as "federal rules of decision," City Brief 15, is

"a destructive oversimplification of the highly intricate interplay of the States and the National Government in their regulation of maritime commerce. It is true that state law must yield to the needs of a uniform federal maritime law when this Court finds inroads on a harmonious system. But this limitation still leaves the States a wide scope." Romero v. International Terminal Operating Co., 358 U.S. 354, 373 (1959) (footnote omitted).

Finally, on top of these objections going to the city's premises there is added a most powerful one based on the practical consequences of adopting a multifactor test. Although the existing case law tempers the locality test with the added requirements looking to potential harm and traditional activity, it reflects customary practice in seeing jurisdiction as the norm when the tort originates with a vessel in navigable waters, and in treating departure from the locality principle as the exception. For better or worse, the case law has thus carved out the approximate shape of admiralty jurisdiction in a way that admiralty lawyers understand reasonably well. As against this approach, so familiar and relatively easy, the proposed four- or seven-factor test would be hard to apply, jettisoning relative predictability for the open-ended rough-and-tumble of factors, inviting complex argument in a trial court and a virtually inevitable appeal.

Consider, for example, just one of the factors under the city's test, requiring a district court at the beginning of every purported admiralty case to determine the source (state or federal) of the applicable substantive law. The difficulty of doing that was an important reason why this Court in Romero, supra, was unable to hold that maritime claims fell within the scope of the federal-question-jurisdiction statute, 28 U.S.C. 1331. 358 U.S., at 375 -376 ("sound judicial policy does not encourage a situation which necessitates constant adjudication of the boundaries of state and federal competence"). That concern applies just as strongly to cases invoking a district court's admiralty jurisdiction under 28 U.S.C. 1333, under which the jurisdictional enquiry for maritime torts has traditionally been quite uncomplicated.

Reasons of practice, then, are as weighty as reasons of theory for rejecting the city's call to adopt a multifactor test for admiralty jurisdiction for the benefit of land-based parties to a tort action.

Accordingly, we conclude that the Court of Appeals correctly held that the District Court had admiralty jurisdiction over the respondent's Limitation Act suit. The judgment of the Court of Appeals is

Affirmed.

Footnotes

1. Grubart argues, based on Margin v. Sea-Land Services, Inc., 812 F.2d 973, 975 (CA5 1987), that an appurtenance is considered part of the vessel only when it is defective. See Brief for Petitioner in No. 93-762, pp. 34-35 (Grubart Brief). Margin, however, does not so hold. It dealt with a land based crane that lowered a ship's hatch cover dangerously close to a welder working on a dock, and its result turned not on the condition of the hatch cover, the putative appurtenance, but on the fact that the plaintiff did not allege that "vessel negligence proximately caused his injury." 812 F.2d, at 977. Indeed, the argument that Congress intended admiralty jurisdiction to extend to injuries caused by defective appurtenances, but not to appurtenances in good condition when operated negligently, makes no sense. See Gutierrez, 373 U.S., at 210 ("There is no distinction in admiralty between torts committed by the ship itself and by the ship's personnel while operating it . . . .").

2. At oral argument, counsel for the city undercut this argument by conceding that admiralty jurisdiction would govern claims arising from an incident in which a ship on navigable waters slipped its moorings, drifted into a dam, and caused a breach in the dam that resulted in flooding of surrounding territory. Tr. of Oral Arg. 17.

3. The city in part bases its assertion about the practical effects of a proximate cause rule on a reading of Crowell v. Benson, 285 U.S. 22, 54-56 (1932), which, according to the city, held that the Longshoremen's and Harbor Worker's Compensation Act could not constitutionally apply to an employee absent a finding that he was actually injured on navigable waters. Thus, the city argues, a construction of the Extension Act that would permit the assertion of federal jurisdiction over land-based injuries absent a finding, on the merits, of actual causation "would raise serious constitutional questions." See Brief for Petitioner in No. 93-1094, pp. 41-42 (City Brief).

Even if the city's interpretation of Crowell is correct, it is not dispositive here. Constitutional difficulties need not arise when a court defers final determination of facts upon which jurisdiction depends until after the first jurisdictional skirmish. In the standing context, for example, we have held that "the Constitution does not require that the plaintiff offer . . . proof [of the facts showing that the plaintiff sustained actual injury] as a threshold matter in order to invoke the District Court's jurisdiction." Gwaltney of Smithfield, Ltd. v. Chesapeake Bay Foundation, Inc., 484 U.S. 49, 66 (1987). We see no reason why a different rule should apply here, and find ourselves in the company of the city's own amici. See Brief for National Conference of State Legislatures et al. as Amici Curiae 18-19, n. 9 (suggesting that "a court need not decide the merits of causation issues to resolve a jurisdictional challenge").

4. The city also proposes that we define the activity as "the operation of an underground tunnel connected to Loop buildings." City Brief 49-50. But doing this would eliminate the maritime tortfeasor's activity from consideration entirely. This (like the choice of a supreme level of generality, described in the text) would turn Sisson v. Ruby, 497 U.S. 358 (1990), on its head, from a test to weed out torts without a maritime connection into an arbitrary exercise for eliminating jurisdiction over even vessel-related torts connected to traditional maritime commerce.

5. Because we conclude that the tort alleged in Count I of Great Lakes's complaint satisfies both the location and connection tests necessary for admiralty jurisdiction under 28 U.S.C. 1333(1), we need not consider respondent's alternative argument that the Extension of Admiralty Jurisdiction Act, 46 U.S.C. App. 740, provides an independent basis of federal jurisdiction over the complaint.

6. We will content ourselves simply with raising a question about another of the city's assumptions, which does not go to anything dispositive for us. It is true that this Court has said that "the primary focus of admiralty jurisdiction is unquestionably the protection of maritime commerce," Foremost Ins. Co. v. Richardson, 457 U.S. 668, 674 (1982); see Sisson, 497 U.S., at 367 ; see id., at 364, n. 2, a premise that recently has been questioned, see Casto, The Origins of Federal Admiralty Jurisdiction in an Age of Privateers, Smugglers, and Pirates, 37 Am. J. Legal Hist. 117 (1993). However that may be, this Court has never limited the interest in question to the "protection of maritime commerce through uniform rules of decision," as the city would have it. City Brief 19. Granted, whatever its precise purpose, it is likely that Congress thought of uniformity of substantive law as a subsidiary goal conducive to furthering that purpose. See Currie, Federalism and the Admiralty: "The Devil's Own Mess," 1960 S. Ct. Rev. 158, 163 ("[A] uniform law was apparently one reason for the establishment of the admiralty jurisdiction in 1789" (footnote omitted)). But we are unwilling to rule out that the first Congress saw a value in federal admiralty courts beyond fostering uniformity of substantive law, stemming, say, from a concern with local bias similar to the presupposition for diversity jurisdiction. See The Federalist No. 80, p. 538 (J. Cooke ed. 1961) (A. Hamilton) ("maritime causes . . . so commonly affect the rights of foreigners"); 1 M. Farrand, Records of the Federal Convention of 1787, p. 124 (1911); 2 id., at 46; see generally D. Robertson, Admiralty and Federalism 95-103 (1970). After all, if uniformity of substantive law had been Congress's only concern, it could have left admiralty jurisdiction in the state courts subject to an appeal to a national tribunal (as it did with federal-question jurisdiction until 1875, and as the Articles of Federation had done with cases of prize and capture).

JUSTICE O'CONNOR, concurring.

I concur in the Court's judgment and opinion. The Court properly holds that, when a court is faced with a case involving multiple tortfeasors, some of whom may not be maritime actors, if one of the putative tortfeasors was engaged in traditional maritime activity alleged to have proximately caused the incident, then the supposedly wrongful activity "involves" traditional maritime activity. The possible involvement of other, nonmaritime parties does not affect the jurisdictional inquiry as to the maritime party. Ante, at 13-14. I do not, however, understand the Court's opinion to suggest that, having found admiralty jurisdiction over a particular claim against a particular party, a court must then exercise admiralty jurisdiction over all the claims and parties involved in the case. Rather, the court should engage in the usual supplemental jurisdiction and impleader inquiries. See 28 U.S.C. 1367 (1988 ed., Supp. V); Fed. Rule Civ. Proc. 14; see also ante, at 3. I find nothing in the Court's opinion to the contrary.

JUSTICE THOMAS, with whom JUSTICE SCALIA joins, concurring in the judgment.

I agree with the majority's conclusion that 28 U.S.C. 1333(1) grants the District Court jurisdiction over the great Chicago flood of 1992. But I write separately because I cannot agree with the test the Court applies to determine the boundaries of admiralty and maritime jurisdiction. Instead of continuing our unquestioning allegiance to the multi-factor approach of Sisson v. Ruby, 497 U.S. 358 (1990), I would restore the jurisdictional inquiry to the simple question whether the tort occurred on a vessel on the navigable waters of the United States. If so, then admiralty jurisdiction exists. This clear, bright-line rule, which the Court applied until recently, ensures that judges and litigants will not waste their resources in determining the extent of federal subject-matter jurisdiction.
 

As recently as 1972, courts and parties experienced little difficulty in determining whether a case triggered admiralty jurisdiction, thanks to the simple "situs rule." In The Plymouth, 3 Wall. 20, 36 (1866), this Court articulated the situs rule thus: "[e]very species of tort, however occurring, and whether on board a vessel or not, if upon the high seas or navigable waters, is of admiralty cognizance." This simple, clear test, test, which Justice Story pronounced while riding circuit, see Thomas v. Lane, 23 F. Cas. 957, 960 (No. 13,902) (CC Me. 1813), did not require alteration until 1948, when Congress included within the admiralty jurisdiction torts caused on water, but whose effects were felt on land. See Extension of Admiralty Jurisdiction Act, 62 Stat. 496, 46 U.S.C. App. 740.

The simplicity of this test was marred by modern cases that tested the boundaries of admiralty jurisdiction with ever more unusual facts. In Executive Jet Aviation, Inc. v. City of Cleveland, 409 U.S. 249 (1972), we held that a plane crash in Lake Erie was not an admiralty case within the meaning of 1333(1) because the tort did not "bear a significant relationship to traditional maritime activity." Id., at 268. What subsequent cases have failed to respect, however, is Executive Jet's clear limitation to torts involving aircraft. As we said:

"One area in which locality as the exclusive test of admiralty tort jurisdiction has given rise to serious problems in application is that of aviation. . . . [W]e have concluded that maritime locality alone is not a sufficient predicate for admiralty jurisdiction in aviation tort cases." Id., at 261 (emphasis added).

Although it modified the strict locality test, Executive Jet still retained a clear rule that I could apply comfortably to the main business of the admiralty court. Nonetheless, the simplicity and clarity of this approach met its demise in Foremost Ins. Co. v. Richardson, 457 U.S. 668 (1982). That case involved the collision of two pleasure boats on the navigable waters, a tort that some commentators had argued did not fall within the admiralty jurisdiction because it did not implicate maritime commerce. See, e.g., Stolz, Pleasure Boating and Admiralty: Erie at Sea, 51 Calif. L. Rev. 661 (1963). The Court could have resolved the case and found jurisdiction simply by applying the situs test. Instead, responding to the arguments that admiralty jurisdiction was limited to commercial maritime activity, the Court found that the tort's "significant connection with traditional maritime activity" and the accident's "potential disruptive impact" on maritime commerce prompted an exercise of federal jurisdiction. 457 U.S., at 674 -675.

It is clear that Foremost overextended Executive Jet, which had reserved the significant relationship inquiry for aviation torts. As JUSTICE SCALIA noted in Sisson, Executive Jet is better "understood as resting on the quite simple ground that the tort did not involve a vessel, which had traditionally been thought required by the leading scholars in the field." 497 U.S., at 369 -370 (opinion concurring in judgment). Executive Jet did not in the least seek to alter the strict locality test for torts involving waterborne vessels. Foremost, however, converted Executive Jet's exception into the rule. In addition to examining situs, Foremost required federal courts to ask whether the tort bore a significant relationship to maritime commerce, and whether the accident had a potential disruptive impact on maritime commerce. 457 U.S., at 673 -675. The lower courts adopted different approaches as they sought to apply Foremost's alteration of the Executive Jet test. See Sisson, supra, at 365, n. 4 (citing cases).

Sisson then affirmed the inherent vagueness of the Foremost test. Sisson involved a marina fire that was caused by a faulty washer-dryer unit on a pleasure yacht. The fire destroyed the yacht and damaged several vessels in addition to the marina. In finding admiralty jurisdiction, the Court held that the federal judicial power would extend to such cases only if: (1) in addition to situs, (2) the "incident" poses a potential hazard to maritime commerce, and (3) the "activity" giving rise to the incident bears a substantial relationship to traditional maritime activity. 497 U.S., at 362 -364. The traditional situs test also would have sustained a finding of jurisdiction because the fire started on board a vessel on the waterways. Thus, what was once a simple question - did the tort occur on the navigable waters - had become a complicated, multi-factor analysis.

The disruption and confusion created by the Foremost-Sisson approach is evident from the post-Sisson decisions of the lower courts and from the majority opinion itself. Faced with the task of determining what is an "incident" or "activity" for Sisson purposes, the Fourth, Fifth, and Ninth Circuits simply reverted to the multi-factor test they had employed before Sisson. See Price v. Price, 929 F.2d 131, 135-136 (CA4 1991); Coats v. Penrod Drilling Corp., 5 F.3d 877, 885-886 (CA5 1993); Delta Country Ventures, Inc. v. Magana, 986 F.2d 1260, 1263 (CA9 1993). The District Court's opinion in this case is typical: while nodding to Sisson, the court focused its entire attention on a totality-of-the-circumstances test, which includes factors such as "the functions and roles of the parties" and "[t]he traditional concepts of the role of admiralty law." Pet. for Cert. of Chicago 32a. Such considerations have no place in the Sisson test and should have no role in any jurisdictional inquiry. The dangers of a totality-of-the-circumstances approach to jurisdiction should be obvious. An undefined test requires courts and litigants to devote substantial resources to determine whether a federal court may hear a specific case. Such a test also introduces undesirable uncertainty into the affairs of private actors - even those involved in common maritime activities - who cannot predict whether or not their conduct may justify the exercise of admiralty jurisdiction.

Although the majority makes an admirable attempt to clarify what Sisson obscures, I am afraid that its analysis cannot mitigate the confusion of the Sisson test. Thus, faced with the "potential to disrupt maritime commerce" prong ante, at 10, the majority must resort to "an intermediate level of possible generality" to determine the "`general features'" of the incident here, id., at 11. The majority does not explain the origins of "levels of generality," nor, to my knowledge, do we employ such a concept in other areas of jurisdiction. We do not use "levels of generality" to characterize residency or amount in controversy for diversity purposes, or to determine the presence of a federal question. Nor does the majority explain why an "intermediate" level of generality is appropriate. It is even unclear what an intermediate level of generality is, and we cannot expect that district courts will apply such a concept uniformly in similar cases. It is far from obvious how the undefined intermediate level of generality indicates that the "incident" for Sisson purposes is that of a vessel damaging an underwater structure.

The majority also applies levels of generality to the next prong of Sisson - whether the tortfeasor is engaged in "activity" that shows a "substantial relationship to traditional maritime activity." The majority decides that the activity is repair work by a vessel on a navigable waterway. But, as the petitioners rightly argue, the "activity" very well could be bridge repair or pile driving. One simply cannot tell due to the ambiguities intrinsic to Sisson and to the uncertainty as to the meaning of levels of generality. The majority's response implicitly acknowledges the vagueness inherent in Sisson: "Although there is inevitably some play in the joints in selecting the right level of generality when applying the Sisson test, the inevitable imprecision is not an excuse for whimsy." Ante, at 14. The Court cannot provide much guidance to district courts as to the correct level of generality; instead, it can only say that any level is probably sufficient so long as it does not lead to "whimsy." When it comes to these issues, I prefer a clearer rule, which this Court has demanded with respect to federal question or diversity jurisdiction. Indeed, the "play in the joints" and "imprecision" that the Court finds "inevitable" easily could be avoided by returning to the test that prevailed before Foremost. In its effort to create an elegant, general test that could include all maritime torts, Sisson has only disrupted what was once a simple inquiry.

In place of Sisson I would follow the test described at the outset. When determining whether maritime jurisdiction exists under 1333(1), a federal district court should ask if the tort occurred on a vessel on the navigable waters. This approach won the approval of two Justices in Sisson, see 497 U.S., at 373 (SCALIA, J., joined by White, J., concurring in judgment). Although JUSTICE SCALIA's Sisson concurrence retained a "normal maritime activities" component, it recognized that anything a vessel does in the navigable waters would meet that requirement, and that "[i]t would be more straightforward to jettison the `traditional maritime activity' analysis entirely." Id., at 374. I wholly agree and have chosen the straightforward approach, which, for all of its simplicity, would have produced the same results the Court arrived at in Executive Jet, Foremost, Sisson, and this case. Although this approach "might leave within admiralty jurisdiction a few unusual actions," ibid., such freakish cases will occur rarely. In any event, the resources needed to resolve them "will be saved many times over by a clear jurisdictional rule that makes it unnecessary to decide" what is a traditional maritime activity and what poses a threat to maritime commerce. Id., at 374-375.

In this case, a straightforward application of the proposed test easily produces a finding of admiralty jurisdiction. As the majority quite ably demonstrates, the situs requirement is satisfied because the tort was caused by a "spud barge" on the Chicago River. Ante, at 6-8. Although the accident's effects were felt on land, the Extension of Admiralty Jurisdiction Act brings the event within 1333(1). While I agree with the majority's analysis of this question, I disagree with its decision to continue on to other issues. A simple application of the situs test would yield the same result the Court reaches at the end of its analysis.

This Court pursues clarity and efficiency in other areas of federal subject-matter jurisdiction, and it should demand no less in admiralty and maritime law. The test I have proposed would produce much the same results as the Sisson analysis without the need for wasteful litigation over threshold jurisdictional questions. Because Sisson departed from a century of precedent, is unworkable, and is easily replaced with a bright-line rule, I concur only in the judgment.