Respondent Latsis' duties as a superintendent engineer for petitioner Chandris, Inc., required him to take voyages on Chandris' ships. He lost substantial vision in one eye after a condition that he developed while on one of those voyages went untreated by a ship's doctor. Following his recuperation, he sailed to Germany on the S. S. Galileo and stayed with the ship while it was in drydock for refurbishment. Subsequently, he sued Chandris for damages for his eye injury under the Jones Act, which provides a negligence cause of action for "any seaman" injured "in the course of his employment." The District Court instructed the jury that Latsis was a "seaman" if he was permanently assigned to, or performed a substantial part of his work on, a vessel, but that the time Latsis spent with the Galileo while it was in drydock could not be considered because the vessel was then out of navigation. The jury returned a verdict for Chandris based solely on Latsis' seaman status. The Court of Appeals vacated the judgment, finding that the jury instruction improperly framed the issue primarily in terms of Latsis' temporal relationship to the vessel. It held that the "employment-related connection to a vessel in navigation" required for seaman status under the Jones Act, McDermott International, Inc. v. Wilander, 498 U.S. 337, 355 , exists where an individual contributes to a vessel's function or the accomplishment of its mission; the contribution is limited to a particular vessel or identifiable group of vessels; the contribution is substantial in terms of its duration or nature; and the course of the individual's employment regularly exposes him to the hazards of the sea. It also found that the District Court erred in instructing the jury that the Galileo's drydock time could not count in the substantial connection equation.

Held:

1. The "employment-related connection to a vessel in navigation" necessary for seaman status comprises two basic elements: the worker's duties must contribute to the function of the vessel or to the accomplishment of its mission, id., at 355, and the worker must have a connection to a vessel in navigation (or an identifiable group of vessels) that is substantial in both its duration and its nature. Pp. 6-25.

(a) The Jones Act provides heightened legal protections to seamen because of their exposure to the perils of the sea, but does not define the term "seaman." However, the Court's Jones Act cases establish the basic principles that the term does not include land-based workers, id., at 348, and that seaman status depends "not on the place where the injury is inflicted . . . but on the nature of the seaman's service, his status as a member of the vessel, and his relationship . . . to the vessel and its operation in navigable waters," Swanson v. Marra Brothers, Inc., 328 U.S. 1, 4 . Thus, land-based maritime workers do not become seamen when they happen to be working aboard a vessel, and seamen do not lose Jones Act coverage when their service to a vessel takes them ashore. Latsis' proposed "voyage test" - under which any maritime worker assigned to a vessel for the duration of a voyage, whose duties contribute to the vessel's mission, would be a seaman for injuries incurred during that voyage - conflicts with this status-based inquiry. Desper v. Starved Rock Ferry Co., 342 U.S. 187, 190 , and Grimes v. Raymond Concrete Pile Co, 356 U.S. 252, 255 , distinguished. Pp. 6-16.

(b) Beyond the basic themes outlined here, the Court's cases have been silent as to the precise relationship a maritime worker must bear to a vessel in order to come within the Jones Act's ambit, leaving the lower federal courts the task of developing appropriate criteria to distinguish "ship's company" from land-based maritime workers. Those courts generally require at least a significant connection to a vessel in navigation (or to an identifiable fleet of vessels) for a maritime worker to qualify as a seaman under the Jones Act. Pp. 16-20.

(c) The test for seaman status adopted here has two essential requirements. The first is a broad threshold requirement that makes all maritime employees who do the ship's work eligible for seaman status. Wilander, supra, at 355. The second requirement determines which of these eligible maritime employees have the required employment-related connection to a vessel in navigation to make them in fact entitled to Jones Act benefits. This requirement gives full effect to the remedial scheme created by Congress and separates sea-based maritime employees entitled to Jones Act Page III protection from land-based workers whose employment does not regularly expose them to the perils of the sea. Who is a "member of a crew" is a mixed question of law and fact. A jury should be able to consider all relevant circumstances bearing on the two requirements. The duration of a worker's connection to a vessel and the nature of the worker's activities, taken together, determine whether he is a seaman, because the ultimate inquiry is whether the worker is part of the vessel's crew or simply a land-based employee who happens to be working on the vessel at a given time. Although seaman status is not merely a temporal concept, it includes a temporal element. A worker who spends only a small fraction of his working time aboard a vessel is fundamentally land-based and therefore not a crew member regardless of his duties. An appropriate rule of thumb is that a worker who spends less than about 30 percent of his time in the service of a vessel in navigation should not qualify as a seaman. This figure is only a guideline that allows a court to take the question from the jury when a worker has a clearly inadequate temporal connection to the vessel. On the other hand, the seaman status inquiry should not be limited exclusively to an examination of the overall course of a worker's service with a particular employer, since his seaman status may change with his basic assignment. Pp. 20-25.

2. The District Court's drydock instruction was erroneous. Whether a vessel is in navigation is a fact-intensive question that can be removed from the jury's consideration only where the facts and the law will reasonably support one conclusion. Based upon the record here, the trial court failed adequately to justify its decision to remove that question from the jury. Moreover, the court's charge to the jury swept too broadly in prohibiting the jury from considering the time Latsis spent with the vessel while in drydock for any purpose. Pp. 25-29.

O'CONNOR, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and SCALIA, KENNEDY, SOUTER, and GINSBURG, JJ., joined. STEVENS, J., filed an opinion concurring in the judgment, in which THOMAS and BREYER, JJ., joined.

JUSTICE O'CONNOR delivered the opinion of the Court.

This case asks us to clarify what "employment-related connection to a vessel in navigation," McDermott International, Inc. v. Wilander, 498 U.S. 337, 355 (1991), is necessary for a maritime worker to qualify as a seaman under the Jones Act, 46 U.S.C. App. 688(a). In Wilander, we addressed the type of activities that a seaman must perform and held that, under the Jones Act, a seaman's job need not be limited to transportation-related functions that directly aid in the vessel's navigation. We now determine what relationship a worker must have to the vessel, regardless of the specific tasks the worker undertakes, in order to obtain seaman status.

On May 14, 1989, Latsis sailed for Bermuda aboard the S. S. Galileo to plan for an upcoming renovation of the ship, which was one of the older vessels in the Chandris fleet. Latsis developed a problem with his right eye on the day of departure, and he saw the ship's doctor as the Galileo left port. The doctor diagnosed a suspected detached retina but failed to follow standard medical procedure, which would have been to direct Latsis to see an ophthalmologist on an emergency basis. Instead, the ship's doctor recommended that Latsis relax until he could see an eye specialist when the Galileo arrived in Bermuda two days later. No attempt was made to transport Latsis ashore for prompt medical care by means of a pilot vessel or helicopter during the 11 hours it took the ship to reach the open sea from Baltimore, and Latsis received no further medical care until after the ship arrived in Bermuda. In Bermuda, a doctor diagnosed a detached retina and recommended immediate hospitalization and surgery. Although the operation was a partial success, Latsis lost 75 percent of his vision in his right eye.

Following his recuperation, which lasted approximately six weeks, Latsis resumed his duties with Chandris. On September 30, 1989, he sailed with the Galileo to Bremerhaven, Germany, where the vessel was placed in drydock for a 6-month refurbishment. After the conversion, the company renamed the vessel the S. S. Meridian. Latsis, who had been with the ship the entire time it was in drydock in Bremerhaven, sailed back to the United States on board the Meridian and continued to work for Chandris until November 1990, when his employment was terminated for reasons that are not clear from the record.

In October 1991, Latsis filed suit in the United States District Court for the Southern District of New York seeking compensatory damages under the Jones Act, 46 U.S.C. App. 688, for the negligence of the ship's doctor that resulted in the significant loss of sight in Latsis' right eye. The Jones Act provides, in pertinent part, that "[a]ny seaman who shall suffer personal injury in the course of his employment may, at his election, maintain an action for damages at law, with the right of trial by jury . . . ." The District Court instructed the jury that it could conclude that Latsis was a seaman within the meaning of the statute if it found as follows:

"[T]he plaintiff was either permanently assigned to the vessel or performed a substantial part of his work on the vessel. In determining whether Mr. Latsis performed a substantial part of his work on the vessel, you may not consider the period of time the Galileo was in drydock in Germany, because during that time period she was out of navigation. You may, however, consider the time spent sailing to and from Germany for the conversion. Also, on this first element of being a seaman, seamen do not include land-based workers." App. 210.

Respondent appealed to the Court of Appeals for the Second Circuit, which vacated the judgment and remanded the case for a new trial. 20 F.3d 45 (1994). The court emphasized that its longstanding test for seaman status under the Jones Act required "`a more or less permanent connection with the ship,'" Salgado v. M. J. Rudolph Corp., 514 F.2d 750, 755 (CA2 1975), a connection that need not be limited to time spent on the vessel but could also be established by the nature of the work performed. The court thought that the alternate formulation employed by the District Court (permanent assignment to the vessel or performance of a substantial part of his work on the vessel), which was derived from Offshore Co. v. Robison, 266 F.2d 769, 779 (CA5 1959), improperly framed the issue for the jury primarily, if not solely, in terms of Latsis' temporal relationship to the vessel. With that understanding of what the language of the Robison test implied, the court concluded that the District Court's seaman status jury instructions constituted plain error under established circuit precedent. The court then took this case as an opportunity to clarify its seaman status requirements, directing the District Court that the jury should be instructed on remand as follows:

"[T]he test of seaman status under the Jones Act is an employment-related connection to a vessel in navigation. The test will be met where a jury finds that (1) the plaintiff contributed to the function of or helped accomplish the mission of, a vessel; (2) the plaintiff's contribution was limited to a particular vessel or identifiable group of vessels; (3) the plaintiff's contribution was substantial in terms of its (a) duration or (b) nature; and (4) the course of the plaintiff's employment regularly exposed the plaintiff to the hazards of the sea." 20 F.3d, at 57.

We granted certiorari, 513 U.S. ___ (1994), to resolve the continuing conflict among the Courts of Appeals regarding the appropriate requirements for seaman status under the Jones Act.*
 

The Jones Act, however, does not define the term "seaman" and therefore leaves to the courts the determination of exactly which maritime workers are entitled to admiralty's special protection. Early on, we concluded that Congress intended the term to have its established meaning under the general maritime law at the time the Jones Act was enacted. See Warner v. Goltra, 293 U.S. 155, 159 (1934). In Warner, we stated that "a seaman is a mariner of any degree, one who lives his life upon the sea." Id., at 157. Similarly, in Norton v. Warner Co., 321 U.S. 565, 572 (1944), we suggested that "`every one is entitled to the privilege of a seaman who, like seamen, at all times contributes to the labors about the operation and welfare of the ship when she is upon a voyage'" (quoting The Buena Ventura, 243 F. 797, 799 (SDNY 1916)).

Congress provided some content for the Jones Act requirement in 1927 when it enacted the Longshore and Harbor Workers' Compensation Act (LHWCA), which provides scheduled compensation (and the exclusive remedy) for injury to a broad range of land-based maritime workers but which also explicitly excludes from its coverage "a master or member of a crew of any vessel." 44 Stat. (part 2) 1424, as amended, 33 U.S.C. 902(3)(G). As the Court has stated on several occasions, the Jones Act and the LHWCA are mutually exclusive compensation regimes: "`master or member of a crew' is a refinement of the term `seaman' in the Jones Act; it excludes from LHWCA coverage those properly covered under the Jones Act." Wilander, 498 U.S., at 347 . Indeed, "it is odd but true that the key requirement for Jones Act coverage now appears in another statute." Ibid. Injured workers who fall under neither category may still recover under an applicable state workers' compensation scheme or, in admiralty, under general maritime tort principles (which are admittedly less generous than the Jones Act's protections). See Cheavens, Terminal Workers' Injury and Death Claims, 64 Tulane L. Rev. 361, 364-365 (1989).

Despite the LHWCA language, drawing the distinction between those maritime workers who should qualify as seamen and those who should not has proved to be a difficult task and the source of much litigation particularly because "the myriad circumstances in which men go upon the water confront courts not with discrete classes of maritime employees, but rather with a spectrum ranging from the blue-water seaman to the land-based longshoreman." Brown v. ITT Rayonier, Inc., 497 F.2d 234, 236 (CA5 1974). The federal courts have struggled over the years to articulate generally applicable criteria to distinguish among the many varieties of maritime workers, often developing detailed multi-pronged tests for seaman status. Since the 1950s, this Court largely has left definition of the Jones Act's scope to the lower courts. Unfortunately, as a result, "[t]he perils of the sea, which mariners suffer and shipowners insure against, have met their match in the perils of judicial review." Gilmore & Black, supra, 6-1, at 272. Or, as one court paraphrased Diderot in reference to this body of law: "`We have made a labyrinth and got lost in it. We must find our way out.'" Johnson v. John F. Beasley Constr. Co., 742 F.2d 1054, 1060 (CA7 1984), cert. denied, 469 U.S. 1211 (1985); see 9 Diderot, Oeuvres Completes 203 (J. Assezat ed. 1875).

After surveying the history of an "aid in navigation" requirement under both the Jones Act and general maritime law, we concluded that "all those with that `peculiar relationship to the vessel' are covered under the Jones Act, regardless of the particular job they perform," 498 U.S., at 354 , and that "the better rule is to define `master or member of a crew' under the LHWCA, and therefore `seaman' under the Jones Act, solely in terms of the employee's connection to a vessel in navigation," ibid. Thus, we held that, although "[i]t is not necessary that a seaman aid in navigation or contribute to the transportation of the vessel, . . . a seaman must be doing the ship's work." Id., at 355. We explained that "[t]he key to seaman status is employment-related connection to a vessel in navigation," and that, although "[w]e are not called upon here to define this connection in all details, . . . we believe the requirement that an employee's duties must `contribut[e] to the function of the vessel or to the accomplishment of its mission' captures well an important requirement of seaman status." Ibid.

Beyond dispensing with the "aid to navigation" requirement, however, Wilander did not consider the requisite connection to a vessel in any detail and therefore failed to end the prevailing confusion regarding seaman status.

Our Jones Act cases establish several basic principles regarding the definition of a seaman. First, "[w]hether under the Jones Act or general maritime law, seamen do not include land-based workers." Wilander, supra, at 348; see also Allbritton, Seaman Status in Wilander's Wake, 68 Tulane L. Rev. 373, 387 (1994). Our early Jones Act decisions had not recognized this fundamental distinction. In International Stevedoring Co. v. Haverty, 272 U.S. 50 (1926), we held that a longshoreman injured while stowing cargo, and while aboard but not employed by a vessel at dock in navigable waters, was a seaman covered by the Jones Act. Recognizing that "for most purposes, as the word is commonly used, stevedores are not `seamen,'" the Court nevertheless concluded that "[w]e cannot believe that Congress willingly would have allowed the protection to men engaged upon the same maritime duties to vary with the accident of their being employed by a stevedore rather than by the ship." Id., at 52. Because stevedores are engaged in "a maritime service formerly rendered by the ship's crew," ibid. (citing Atlantic Transport Co. of W. Va. v. Imbrovek, 234 U.S. 52, 62 (1914)), we concluded, they should receive the Jones Act's protections. See also Uravic v. F. Jarka Co., 282 U.S. 234, 238 (1931); Jamison v. Encarnacion, 281 U.S. 635, 639 (1930). In 1946, the Court belatedly recognized that Congress had acted, in passing the LHWCA in 1927, to undercut the Court's reasoning in the Haverty line of cases and to emphasize that land-based maritime workers should not be entitled to the seamen's traditional remedies. Our decision in Swanson v. Marra Brothers, Inc., 328 U.S. 1, 7 (1946), acknowledged that Congress had expressed its intention to "confine the benefits of the Jones Act to the members of the crew of a vessel plying in navigable waters and to substitute for the right of recovery recognized by the Haverty case only such rights to compensation as are given by [the LHWCA]." See also South Chicago Coal & Dock Co. v. Bassett, 309 U.S. 251, 257 (1940). Through the LHWCA, therefore, Congress "explicitly den[ied] a right of recovery under the Jones Act to maritime workers not members of a crew who are injured on board a vessel." Swanson, supra, at 6. And this recognition process culminated in Wilander with the Court's statement that, "[w]ith the passage of the LHWCA, Congress established a clear distinction between land-based and sea-based maritime workers. The latter, who owe their allegiance to a vessel and not solely to a land-based employer, are seamen." 498 U.S., at 347 .

In addition to recognizing a fundamental distinction between land-based and sea-based maritime employees, our cases also emphasize that Jones Act coverage, like the jurisdiction of admiralty over causes of action for maintenance and cure for injuries received in the course of a seamen's employment, depends "not on the place where the injury is inflicted . . . but on the nature of the seaman's service, his status as a member of the vessel, and his relationship as such to the vessel and its operation in navigable waters." Swanson, supra, at 4. Thus, maritime workers who obtain seaman status do not lose that protection automatically when on shore and may recover under the Jones Act whenever they are injured in the service of a vessel, regardless of whether the injury occurs on or off the ship. In O'Donnell v. Great Lakes Dredge & Dock Co., 318 U.S. 36 (1943), the Court held a shipowner liable for injuries caused to a seaman by a fellow crew member while the former was on shore repairing a conduit that was a part of the vessel and that was used for discharging the ship's cargo. We explained: "The right of recovery in the Jones Act is given to the seaman as such, and, as in the case of maintenance and cure, the admiralty jurisdiction over the suit depends not on the place where the injury is inflicted but on the nature of the service and its relationship to the operation of the vessel plying in navigable waters." Id., at 42-43. Similarly, the Court in Swanson emphasized that the LHWCA "leaves unaffected the rights of members of the crew of a vessel to recover under the Jones Act when injured while pursuing their maritime employment whether on board . . . or on shore." 328 U.S., at 7 -8. See also Braen v. Pfeiffer Transportation Co., 361 U.S. 129, 131 -132 (1959).

Our LHWCA cases also recognize the converse: land-based maritime workers injured while on a vessel in navigation remain covered by the LHWCA, which expressly provides compensation for injuries to certain workers engaged in "maritime employment" that are incurred "upon the navigable waters of the United States," 33 U.S.C. 903(a). Thus, in Director, OWCP v. Perini North River Associates, 459 U.S. 297 (1983), we held that a worker injured while "working on a barge in actual navigable waters" of the Hudson River, id., at 300, n. 4, could be compensated under the LHWCA, id., at 324. See also Parker v. Motor Boat Sales, Inc., 314 U.S. 244, 244 -245 (1941) (upholding LHWCA coverage for a worker testing outboard motors who "was drowned when a motor boat in which he was riding capsized"). These decisions, which reflect our longstanding view of the LHWCA's scope, indicate that a maritime worker does not become a "member of a crew" as soon as a vessel leaves the dock.

It is therefore well settled after decades of judicial interpretation that the Jones Act inquiry is fundamentally status-based: land-based maritime workers do not become seamen because they happen to be working on board a vessel when they are injured, and seamen do not lose Jones Act protection when the course of their service to a vessel takes them ashore. In spite of this background, respondent and JUSTICE STEVENS suggest that any maritime worker who is assigned to a vessel for the duration of a voyage - and whose duties contribute to the vessel's mission - should be classified as a seaman for purposes of injuries incurred during that voyage. See Brief for Respondent 14; post, at 1 (STEVENS, J., concurring in the judgment). Under such a "voyage test," which relies principally upon this Court's statements that the Jones Act was designed to protect maritime workers who are exposed to the "special hazards" and "particular perils" characteristic of work on vessels at sea, see, e.g., Wilander, supra, at 354, the worker's activities at the time of the injury would be controlling.

The difficulty with respondent's argument, as the foregoing discussion makes clear, is that the LHWCA repudiated the Haverty line of cases and established that a worker is no longer considered to be a seaman simply because he is doing a seaman's work at the time of the injury. Seaman status is not coextensive with seamen's risks. See, e.g., Easley v. Southern Shipbuilding Corp., 965 F.2d 1, 4-5 (CA5 1992), cert. denied, 506 U.S. ___ (1993); Robertson 93 (following "the overwhelming weight of authority in taking it as given that seaman status cannot be established by any worker who fails to demonstrate that a significant portion of his work was done aboard a vessel" and acknowledging that "[s]ome workers who unmistakably confront the perils of the sea, often in extreme form, are thereby left out of the seamen's protections" (footnote omitted)). A "voyage test" would conflict with our prior understanding of the Jones Act as fundamentally status-based, granting the negligence cause of action to those maritime workers who form the ship's company. Swanson, supra, at 4-5; O'Donnell, supra, at 42-43.

Desper v. Starved Rock Ferry Co., 342 U.S. 187, 190 (1952), is not to the contrary. Although some language in that case does suggest that whether an individual is a seaman depends upon "the activity in which he was engaged at the time of injury," the context of that discussion reveals that "activity" referred to the worker's employment as a laborer on a vessel undergoing seasonal repairs while out of navigation, and not to his precise task at the time of injury. Similarly, despite Justice Harlan's suggestion in dissent that the Court's decision in Grimes v. Raymond Concrete Pile Co. necessarily construed the word seaman "to mean nothing more than a person injured while working at sea," 356 U.S. 252, 255 (1958), our short per curiam opinion in that case does not indicate that we adopted so expansive a reading of the statutory term. Citing our prior cases which emphasized that the question of seaman status is normally for the fact finder to decide, see, e.g., Senko v. La Crosse Dredging Corp., 352 U.S. 370, 371 -372 (1957); Bassett, supra, at 257-258, we reversed the judgment of the Court of Appeals and held simply that the jury could have inferred from the facts presented that the petitioner was a member of a crew in light of his overall service to the company (as the District Court had concluded in ruling on a motion for a directed verdict at the close of petitioner's case). Grimes, 356 U.S., at 253. That neither Desper nor Grimes altered our established course in favor of a voyage test is confirmed by reference to our later decision in Braen, 361 U.S., at 131 , in which we repeated that "[t]he injured party must of course have `status as a member of the vessel' for it is seamen, not others who may work on the vessel (Swanson v. Marra Bros., 328 U.S. 1, 4 ), to whom Congress extended the protection of the Jones Act."

We believe it is important to avoid "`engrafting upon the statutory classification of a "seaman" a judicial gloss so protean, elusive, or arbitrary as to permit a worker to walk into and out of coverage in the course of his regular duties.'" Barrett v. Chevron, U.S.A., Inc., 781 F.2d 1067, 1075 (CA5 1986) (en banc) (quoting Longmire v. Sea Drilling Corp., 610 F.2d 1342, 1347, n. 6 (CA5 1980)). In evaluating the employment-related connection of a maritime worker to a vessel in navigation, courts should not employ "a `snapshot' test for seaman status, inspecting only the situation as it exists at the instant of injury; a more enduring relationship is contemplated in the jurisprudence." Easley, supra, at 5. Thus, a worker may not oscillate back and forth between Jones Act coverage and other remedies depending on the activity in which the worker was engaged while injured. Reeves v. Mobile Dredging & Pumping Co., 26 F.3d 1247, 1256 (CA3 1994). Unlike JUSTICE STEVENS, see post, at 8, we do not believe that any maritime worker on a ship at sea as part of his employment is automatically a member of the crew of the vessel within the meaning of the statutory terms. Our rejection of the voyage test is also consistent with the interests of employers and maritime workers alike in being able to predict who will be covered by the Jones Act (and, perhaps more importantly for purposes of the employers' workers' compensation obligations, who will be covered by the LHWCA) before a particular work day begins.

To say that our cases have recognized a distinction between land-based and sea-based maritime workers that precludes application of a voyage test for seaman status, however, is not to say that a maritime employee must work only on board a vessel to qualify as a seaman under the Jones Act. In Southwest Marine, Inc. v. Gizoni, 502 U.S. 81 (1991), decided only a few months after Wilander, we concluded that a worker's status as a ship repairman, one of the enumerated occupations encompassed within the term "employee" under the LHWCA, 33 U.S.C. 902(3), did not necessarily restrict the worker to a remedy under that statute. We explained that, "[w]hile in some cases a ship repairman may lack the requisite connection to a vessel in navigation to qualify for seaman status, . . . not all ship repairmen lack the requisite connection as a matter of law. This is so because `[i]t is not the employee's particular job that is determinative, but the employee's connection to a vessel.'" Gizoni, supra, at 89 (quoting Wilander, 498 U.S., at 354 ) (footnote omitted). Thus, we concluded, the Jones Act remedy may be available to maritime workers who are employed by a shipyard and who spend a portion of their time working on shore but spend the rest of their time at sea.

Beyond these basic themes, which are sufficient to foreclose respondent's principal argument, our cases are largely silent as to the precise relationship a maritime worker must bear to a vessel in order to come within the Jones Act's ambit. We have, until now, left to the lower federal courts the task of developing appropriate criteria to distinguish the "ship's company" from those members of the maritime community whose employment is essentially land-based.

The second major body of seaman status law developed in the Court of Appeals for the Fifth Circuit, which has a substantial Jones Act caseload, in the wake of Offshore Co. v. Robison, 266 F.2d 769 (CA5 1959). At the time of his injury, Robison was an oil worker permanently assigned to a drilling rig mounted on a barge in the Gulf of Mexico. In sustaining the jury's award of damages to Robison under the Jones Act, the court abandoned the aid in navigation requirement of the traditional test and held as follows:

"[T]here is an evidentiary basis for a Jones Act case to go to the jury: (1) if there is evidence that the injured workman was assigned permanently to a vessel . . . or performed a substantial part of his work on the vessel; and (2) if the capacity in which he was employed or the duties which he performed contributed to the function of the vessel or the accomplishment of its mission, or to the operation or welfare of the vessel in terms of its maintenance during its movement or during anchorage for its future trips." Id., at 779 (footnote omitted).

While the Carumbo and Robison approaches may not seem all that different at first glance, subsequent developments in the Fifth Circuit's Jones Act jurisprudence added a strictly temporal gloss to the Jones Act inquiry. Under Barrett v. Chevron, U.S.A., Inc., supra, if an employee's regular duties require him to divide his time between vessel and land, his status as a crew member is determined "in the context of his entire employment" with his current employer. Id., at 1075. See also Allbritton, 68 Tulane L. Rev., at 386; Longmire, 610 F.2d, at 1347 (explaining that a worker's seaman status "should be addressed with reference to the nature and location of his occupation taken as a whole"). In Barrett, the court noted that the worker "performed seventy to eighty percent of his work on platforms and no more than twenty to thirty percent of his work on vessels" and then concluded that, "[b]ecause he did not perform a substantial portion of his work aboard a vessel or fleet of vessels, he failed to establish that he was a member of the crew of a vessel." 781 F.2d, at 1076. Since Barrett, the Fifth Circuit consistently has analyzed the problem in terms of the percentage of work performed on vessels for the employer in question - and has declined to find seaman status where the employee spent less than 30 percent of his time aboard ship. See, e.g., Palmer v. Fayard Moving & Transp. Corp., 930 F.2d 437, 439 (CA5 1991); Lormand v. Superior Oil Co., 845 F.2d 536, 541 (CA5 1987), cert. denied, 484 U.S. 1031 (1988); cf. Leonard v. Dixie Well Service & Supply, Inc., 828 F.2d 291, 295 (CA5 1987); Pickle v. International Oilfield Divers, Inc., 791 F.2d 1237, 1240 (CA5 1986), cert. denied, 479 U.S. 1059 (1987).

Although some courts of appeals have varied the applicable tests to some degree, see, e.g., Johnson v. John F. Beasley Constr. Co., 742 F.2d, at 1062-1063, the traditional Carumbo seaman status formulation and the subsequent Robison modification are universally recognized, and one or the other is applied in every federal circuit to have considered the issue. See Bull, Seaman Status Revisited: A Practical Guide To Status Determination, 6 U.S. F. Mar. L.J. 547, 562-572 (1994) (collecting cases). The federal courts generally require at least a significant connection to a vessel in navigation (or to an identifiable fleet of vessels) for a maritime worker to qualify as a seaman under the Jones Act. Although the traditional test requires a "more or less permanent connection" and the Robison formulation calls for "substantial" work aboard a vessel, "this general requirement varies little, if at all, from one jurisdiction to another," id., at 587, and "[t]he courts have repeatedly held that the relationship creating seaman status must be substantial in point of time and work, and not merely sporadic," id., at 587-588.

Second, and most important for our purposes here, a seaman must have a connection to a vessel in navigation (or to an identifiable group of such vessels) that is substantial in terms of both its duration and its nature. The fundamental purpose of this substantial connection requirement is to give full effect to the remedial scheme created by Congress and to separate the sea-based maritime employees who are entitled to Jones Act protection from those land-based workers who have only a transitory or sporadic connection to a vessel in navigation, and therefore whose employment does not regularly expose them to the perils of the sea. See 1B A. Jenner, Benedict on Admiralty, 11a, pp. 2-10.1 to 2-11 (7th ed. 1994) ("If it can be shown that the employee performed a significant part of his work on board the vessel on which he was injured, with at least some degree of regularity and continuity, the test for seaman status will be satisfied" (footnote omitted)). This requirement therefore determines which maritime employees in Wilander's broad category of persons eligible for seaman status because they are "doing the ship's work," 498 U.S., at 355 , are in fact entitled to the benefits conferred upon seamen by the Jones Act because they have the requisite employment-related connection to a vessel in navigation.

It is important to recall that the question of who is a "member of a crew," and therefore who is a "seaman," is a mixed question of law and fact. Because statutory terms are at issue, their interpretation is a question of law and it is the court's duty to define the appropriate standard. Wilander, 498 U.S., at 356 . On the other hand, "[i]f reasonable persons, applying the proper legal standard, could differ as to whether the employee was a `member of a crew,' it is a question for the jury." Ibid. See also Senko, 352 U.S., at 374 (explaining that "the determination of whether an injured person was a `member of a crew' is to be left to the finder of fact" and that "a jury's decision is final if it has a reasonable basis"). The jury should be permitted, when determining whether a maritime employee has the requisite employment-related connection to a vessel in navigation to qualify as a member of the vessel's crew, to consider all relevant circumstances bearing on the two elements outlined above.

In defining the prerequisites for Jones Act coverage, we think it preferable to focus upon the essence of what it means to be a seaman and to eschew the temptation to create detailed tests to effectuate the congressional purpose, tests that tend to become ends in and of themselves. The principal formulations employed by the Courts of Appeals - "more or less permanent assignment" or "connection to a vessel that is substantial in terms of its duration and nature" - are simply different ways of getting at the same basic point: the Jones Act remedy is reserved for sea-based maritime employees whose work regularly exposes them to "the special hazards and disadvantages to which they who go down to sea in ships are subjected." Sieracki, 328 U.S., at 104 (Stone, C. J., dissenting). Indeed, it is difficult to discern major substantive differences in the language of the two phrases. In our view, "the total circumstances of an individual's employment must be weighed to determine whether he had a sufficient relation to the navigation of vessels and the perils attendant thereon." Wallace v. Oceaneering Int'l, 727 F.2d 427, 432 (CA5 1984). The duration of a worker's connection to a vessel and the nature of the worker's activities, taken together, determine whether a maritime employee is a seaman because the ultimate inquiry is whether the worker in question is a member of the vessel's crew or simply a land-based employee who happens to be working on the vessel at a given time.

Although we adopt the centerpiece of the formulation used by the Court of Appeals in this case - that a seaman must have a connection with a vessel in navigation that is substantial in both duration and nature - we should point out how our understanding of the import of that language may be different in some respects from that of the court below. The Court of Appeals suggested that its test for seaman status "does not unequivocally require a Jones Act seaman to be substantially connected to a vessel" in terms of time if the worker performs important work on board on a steady, although not necessarily on a temporally significant, basis. 20 F.3d, at 53. Perhaps giving effect to this intuition, or perhaps reacting to the temporal gloss placed on the Robison language by later Fifth Circuit decisions, the court phrased its standard at one point as requiring a jury to find that a Jones Act plaintiff's contribution to the function of the vessel was substantial in terms of its duration or nature. Id., at 57. It is not clear which version ("duration or nature" as opposed to "duration and nature") the Court of Appeals intended to adopt for the substantial connection requirement - or indeed whether the court saw a significant difference between the two. Nevertheless, we think it is important that a seaman's connection to a vessel in fact be substantial in both respects.

We agree with the Court of Appeals that seaman status is not merely a temporal concept, but we also believe that it necessarily includes a temporal element. A maritime worker who spends only a small fraction of his working time on board a vessel is fundamentally land-based and therefore not a member of the vessel's crew, regardless of what his duties are. Naturally, substantiality in this context is determined by reference to the period covered by the Jones Act plaintiff's maritime employment, rather than by some absolute measure. Generally, the Fifth Circuit seems to have identified an appropriate rule of thumb for the ordinary case: a worker who spends less than about 30 percent of his time in the service of a vessel in navigation should not qualify as a seaman under the Jones Act. This figure of course serves as no more than a guideline established by years of experience, and departure from it will certainly be justified in appropriate cases. As we have said, "[t]he inquiry into seaman status is of necessity fact specific; it will depend on the nature of the vessel and the employee's precise relation to it." Wilander, 498 U.S., at 356 . Nevertheless, we believe that courts, employers, and maritime workers can all benefit from reference to these general principles. And where undisputed facts reveal that a maritime worker has a clearly inadequate temporal connection to vessels in navigation, the court may take the question from the jury by granting summary judgment or a directed verdict. See, e.g., Palmer, 930 F.2d, at 439.

On the other hand, we see no reason to limit the seaman status inquiry, as petitioners contend, exclusively to an examination of the overall course of a worker's service with a particular employer. Brief for Petitioners 14-15. When a maritime worker's basic assignment changes, his seaman status may change as well. See Barrett, 781 F.2d, at 1077 (Rubin, J., dissenting) ("An assignment to work as a crew member, like the voyage of a vessel, may be brief, and the Robison test is applicable in deciding the worker's status during any such employment"); Longmire, 610 F.2d, at 1347, n. 6. For example, we can imagine situations in which someone who had worked for years in an employer's shoreside headquarters is then reassigned to a ship in a classic seaman's job that involves a regular and continuous, rather than intermittent, commitment of the worker's labor to the function of a vessel. Such a person should not be denied seaman status if injured shortly after the reassignment, just as someone actually transferred to a desk job in the company's office and injured in the hallway should not be entitled to claim seaman status on the basis of prior service at sea. If a maritime employee receives a new work assignment in which his essential duties are changed, he is entitled to have the assessment of the substantiality of his vessel-related work made on the basis of his activities in his new position. See Cheavens, 64 Tulane L. Rev., at 389-390. Thus, nothing in our opinion forecloses Jones Act coverage, in appropriate cases, for Justice Stevens' paradigmatic maritime worker injured while reassigned to "a lengthy voyage on the high seas," post, at 10. While our approach maintains the status-based inquiry this Court's earlier cases contemplate, we recognize that seaman status also should not be some immutable characteristic that maritime workers who spend only a portion of their time at sea can never attain.

The foregoing discussion establishes that, to qualify as a seaman under the Jones Act, a maritime employee must have a substantial employment-related connection to a vessel in navigation. See Wilander, supra, at 354-355. Of course, any time Latsis spent with the Galileo while the ship was out of navigation could not count as time spent at sea for purposes of that inquiry, and it would have been appropriate for the District Court to make this clear to the jury. Yet the underlying inquiry whether a vessel is or is not "in navigation" for Jones Act purposes is a fact-intensive question that is normally for the jury and not the court to decide. See Butler v. Whiteman, 356 U.S. 271 (1958) (per curiam); 2 M. Norris, Law of Seamen, 30.13, p. 363 (4th ed. 1985) ("Whether the vessel is in navigation presents a question of fact to be determined by the trier of the facts. When the case is tried to a jury the fact question should be left to their consideration if sufficient evidence has been presented to provide the basis for jury consideration"). Removing the issue from the jury's consideration is only appropriate where the facts and the law will reasonably support only one conclusion, Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 250 -251 (1986), and the colloquy between the court and counsel does not indicate that the District Court made any such findings before overruling respondent's objection to the drydock instruction. See Tr. 432. Based upon the record before us, we think the court failed adequately to justify its decision to remove the question whether the Galileo was "in navigation" while in Bremerhaven from the jury.

Under our precedent and the law prevailing in the circuits, it is generally accepted that "a vessel does not cease to be a vessel when she is not voyaging, but is at anchor, berthed, or at dockside," DiGiovanni v. Traylor Bros., Inc., 959 F.2d 1119, 1121 (CA1) (en banc), cert. denied, 506 U.S. ___ (1992), even when the vessel is undergoing repairs. See Butler, supra, at 271; Senko, 352 U.S., at 373 ; Norris, supra, at 364 ("[A] vessel is in navigation . . . when it returns from a voyage and is taken to a drydock or shipyard to undergo repairs in preparation to making another trip, and likewise a vessel is in navigation, although moored to a dock, if it remains in readiness for another voyage" (footnotes omitted)). At some point, however, repairs become sufficiently significant that the vessel can no longer be considered in navigation. In West v. United States, 361 U.S. 118 (1959), we held that a shoreside worker was not entitled to recover for unseaworthiness because the vessel on which he was injured was undergoing an overhaul for the purpose of making her seaworthy and therefore had been withdrawn from navigation. We explained that, in such cases, "the focus should be upon the status of the ship, the pattern of the repairs, and the extensive nature of the work contracted to be done." Id., at 122. See also United New York & New Jersey Sandy Hook Pilots Assn. v. Halecki, 358 U.S. 613 (1959); Desper, 342 U.S., at 191 . The general rule among the Courts of Appeals is that vessels undergoing repairs or spending a relatively short period of time in drydock are still considered to be "in navigation" whereas ships being transformed through "major" overhauls or renovations are not. See Bull, 6 U.S. F. Mar. L. J., at 582-584 (collecting cases).

Obviously, while the distinction at issue here is one of degree, the prevailing view is that "major renovations can take a ship out of navigation, even though its use before and after the work will be the same." McKinley v. All Alaskan Seafoods, Inc., 980 F.2d 567, 570 (CA9 1992). Our review of the record in this case uncovered relatively little evidence bearing on the Galileo's status during the repairs, and even less discussion of the question by the District Court. On the one hand, the work on the Chandris vessel took only about six months, which seems to be a relatively short period of time for important repairs on oceangoing vessels. Cf. id., at 571 (17-month-long project involving major structural changes took the vessel out of navigation); Wixom v. Boland Marine & Manufacturing Co., 614 F.2d 956 (CA5 1980) (similar 3-year project); see also Senko, supra, at 373 (noting that "[e]ven a transoceanic liner may be confined to berth for lengthy periods, and while there the ship is kept in repair by its `crew'" - and that "[t]here can be no doubt that a member of its crew would be covered by the Jones Act during this period, even though the ship was never in transit during his employment"). On the other hand, Latsis' own description of the work performed suggests that the modifications to the vessel were actually quite significant, including the removal of the ship's bottom plates and propellers, the addition of bow thrusters, overhaul of the main engines, reconstruction of the boilers, and renovations of the cabins and other passenger areas of the ship. See App. 93-94. On these facts, which are similar to those in McKinley, it is possible that Chandris could be entitled to partial summary judgment or a directed verdict concerning whether the Galileo remained in navigation while in drydock; the record, however, contains no stipulations or findings by the District Court to justify its conclusion that the modifications to the Galileo were sufficiently extensive to remove the vessel from navigation as a matter of law. On that basis, we agree with the Court of Appeals that the District Court's drydock instruction was erroneous.

Even if the District Court had been justified in directing a verdict on the question whether the Galileo remained in navigation while in Bremerhaven, we think that the court's charge to the jury swept too broadly. Instead of simply noting the appropriate legal conclusion and instructing the jury not to consider the time Latsis spent with the vessel in drydock as time spent with a vessel in navigation, the District Court appears to have prohibited the jury from considering Latsis' stay in Bremerhaven for any purpose. In our view, Latsis' activities while the vessel was in drydock are at least marginally relevant to the underlying inquiry (whether Latsis was a seaman and not a land-based maritimeemployee). Naturally, the jury would be free to draw several inferences from Latsis' work during the conversion, not all of which would be in his favor. But the choice among such permissible inferences should have been left to the jury, and we think the District Court's broadly worded instruction improperly deprived the jury of that opportunity by forbidding the consideration of Latsis' time in Bremerhaven at all.

On remand, the District Court should charge the jury in a manner consistent with our holding that the "employment-related connection to a vessel in navigation" necessary to qualify as a seaman under the Jones Act, id., at 355, comprises two basic elements: the worker's duties must contribute to the function of the vessel or to the accomplishment of its mission, and the worker must have a connection to a vessel in navigation (or an identifiable group of vessels) that is substantial in terms of both its duration and its nature. As to the latter point, the court should emphasize that the Jones Act was intended to protect sea-based maritime workers, who owe their allegiance to a vessel, and not land-based employees, who do not. By instructing juries in Jones Act cases accordingly, courts can give proper effect to the remedial scheme Congress has created for injured maritime workers.

It is so ordered.

JUSTICE STEVENS, with whom JUSTICE THOMAS and JUSTICE BREYER join, concurring in the judgment.

The majority has reached the odd conclusion that a maritime engineer, injured aboard ship on the high seas while performing his duties as an employee of the ship, might not be a "seaman" within the meaning of the Jones Act. This decision is unprecedented. It ignores the critical distinction between work performed aboard ship during a voyage - when the members of the crew encounter "the perils of the sea" - and maritime work performed on a vessel moored to a dock in a safe harbor. In my judgment, an employee of the ship who is injured at sea in the course of his employment is always a "seaman." I would leave more ambiguous, shore-bound cases for another day. Accordingly, though I concur in the Court's disposition of this case, returning it to the District Court for a new trial, I disagree with the standard this Court directs the trial court to apply on remand.

This conclusion follows, first, from the language of the Jones Act and of the Longshore and Harbor Workers' Compensation Act (LHWCA), 33 U.S.C. 901, et seq. The latter, a federal workers' compensation scheme for shore-based maritime workers, exempts any "master or member of a crew of any vessel," 33 U.S.C. 902(3) (G) - a formulation that, we have held, is coextensive with the term "seaman" in the Jones Act. McDermott International, Inc. v. Wilander, 498 U.S. 337, 347 (1991). In ordinary parlance, an employee of a ship at sea who is on that ship as part of his employment and who contributes to the ship's mission is both a "seaman" and a "member of [the] crew of [the] vessel." Indeed, I am not sure how these words can reasonably be read to exclude such an employee. Surely none of the statutory language suggests that the individual must be a member of the ship's crew for longer than a single voyage.

My conclusion also comports with the clear purpose of the Jones Act and of the other maritime law remedies traditionally afforded to seamen:3 to protect maritime workers from exposure to the perils of the sea. In Wilander, 498 U.S., at 354 , we endorsed Chief Justice Stone's explanation of the admiralty law's favored treatment of seamen. Chief Justice Stone wrote:

"The liability of the vessel or owner for maintenance and cure, regardless of their negligence, was established long before our modern conception of contract. But it, like the liability to indemnify the seaman for injuries resulting from unseaworthiness, has been universally recognized as an obligation growing out of the status of the seaman and his peculiar relationship to the vessel, and as a feature of the maritime law compensating or offsetting the special hazards and disadvantages to which they who go down to sea in ships are subjected. They are exposed to the perils of the sea and all the risks of unseaworthiness, with little opportunity to avoid those dangers or to discover and protect themselves from them or to prove who is responsible for theunseaworthiness causing the injury.

"For these reasons the seaman has been given a special status in the maritime law as the ward of the admiralty, entitled to special protection of the law not extended to land employees. Justice Story said in Reed v. Canfield, Fed. Cas. No. 11,641, 1 Sumn. 195, 199: `Seamen are in some sort co-adventurers upon the voyage; and lose their wages upon casualties, which do not affect artisans at home. They share the fate of the ship in cases of shipwreck and capture. They are liable to different rules of discipline and sufferings from landsmen. The policy of the maritime law, for great, and wise, and benevolent purposes, has built up peculiar rights, privileges, duties, and liabilities in the sea-service, which do not belong to home pursuits.'" Seas Shipping Co. v. Sieracki, 328 U.S. 85, 104 -105 (1946) (dissenting opinion) (citations omitted).

"It is for these reasons that throughout the long history of the maritime law the right to maintenance and cure, and later the right to indemnity for injuries attributable to unseaworthiness, have been confined to seamen. Longshoremen and harbor workers are in a class very different from seamen, and one not calling for the creation of extraordinary obligations of the vessel or its owner in their favor, more than other classes of essentially land workers. Unlike members of the crew of a vessel they do not go to sea; they are not subject to the rigid discipline of the sea; they are not prevented by law or ship's discipline from leaving the vessel on which they maybe employed; they have the same recourse as land workers to avoid the hazards to which they are exposed, to ascertain the cause of their injury and to prove it in court." Id., at 105.

This conclusion is consistent with every Jones Act case that this Court has decided. Justice Cardozo's opinion for the Court in Warner v. Goltra, 293 U.S. 155 (1934),set a course that we have consistently followed. Explaining our holding that the master of a tugboat is a "seaman," he explained that "[i]t is enough that what he does affects `the operation and welfare of the ship when she is upon a voyage.'" Id., at 157.5   Indeed, apart from the argument that a seaman must assist in performing the transportation function of the vessel - an argument finally put to rest in McDermott International, Inc v. Wilander, 498 U.S. 337 (1991) - I am not aware of a single Jones Act case decided by this Court, other than Warner,6 in which anyone even argued that an employee who was aboard the ship contributing to the ship's mission while the vessel was in navigation on the high seas was not a seaman. In light of the purposes of the Jones Act, that position is simply too farfetched. As a leading admiralty treatise has recognized, "[i]t seems never to have been questioned that any member of a ship's company who actually goes to sea, no matter what his (or her) duties may be, is a seaman." G. Gilmore & C. Black, Law of Admiralty 6-21, p. 331 (2d ed. 1975).

Surely nothing in Wilander contradicts this basic proposition. In that opinion, we made several references to the importance of work performed on a voyage. Thus, we quoted from leading 19th century treatises on admiralty: "`The term mariner includes all persons employed on board ships and vessels during the voyage to assist in their navigation and preservation, or to promote the purposes of the voyage. . . . [A]t all times and in all countries, all the persons who have been necessarily or properly employed in a vessel as co-laborers to the great purpose of the voyage, have, by the law, been clothed with the legal rights of mariners.'" 498 U.S., at 344 -345 (emphasis deleted), quoting E. Benedict, American Admiralty 278, 241, pp. 158, 133-134 (1850). "An 1883 treatise declared: `All persons employed on a vessel to assist in the main purpose of the voyage are mariners, and included under the name of seamen.' M. Cohen, Admiralty 239." 498 U.S., at 346 . Summarizing our conclusion, we wrote:

"We believe the better rule is to define `master or member of a crew' under the LHWCA, and therefore `seaman' under the Jones Act, solely in terms of the employee's connection to a vessel in navigation. This rule best explains our case law and is consistent with the pre-Jones Act interpretation of `seaman' and Congress' land-based/sea-based distinction. All who work at sea in the service of a ship face those particular perils to which the protection of maritime law, statutory as well as decisional, is directed." Id., at 354.

Respondent's argument, "that any worker who is assigned to a vessel for the duration of a voyage and whose duties contribute to the vessel's mission must be classified as a seaman respecting injuries incurred on that voyage," Brief for Respondent 14, is not inconsistent with the Court's view, ante, at 11-13, that an employee must occupy a certain status in order to qualify as a seaman. It merely recognizes that all members of a ship's crew have that status while the vessel is at sea. In contrast, when the ship is in a harbor, further inquiry may be necessary to separate land-based from sea-based maritime employees. The Court is therefore simply wrong when it states that a "`voyage test' would conflict with our prior understanding of the Jones Act as fundamentally status-based, granting the negligence cause of action to those maritime workers who form the ship's company," ante, at 14. The "ship's company" is readily identifiable when the ship is at sea; the fact that it may be less so when the ship is in port is not an acceptable reason for refusing to rely on the voyage test in a case like this one.

The Court is also quite wrong to suggest that our prior cases "indicate that a maritime worker does not become a `member of a crew' as soon as a vessel leaves the dock," ante, at 13. In neither of the two cases on which it relies to support this conclusion did the injured workman even claim the status of a seaman. In Director, Office of Workers' Compensation Programs v. Perini North River Associates, 459 U.S. 297 (1983), we held that an employee of a firm that was building the foundation of a sewage treatment plant, which extended over the Hudson River adjacent to Manhattan, was covered by the LHWCA because he was injured while working on a barge in navigable waters. The Court of Appeals had denied coverage on the ground that this worker was not engaged in maritime employment. Thus, Perini had nothing to do with any possible overlap between the Jones Act and the LHWCA; this Court's reversal merely found a sufficient maritime connection to support LHWCA coverage of an admittedly shore-based worker.

The other case that the Court cites, Parker v. Motor Boat Sales, Inc., 314 U.S. 244 (1941), involved a janitor who had drowned while riding in a motor boat on the James River near Richmond. The Court of Appeals had held that his widow was not entitled to compensation under the LHWCA on the alternative grounds (1) that the janitor was not acting in the course of his employment when the boat capsized, and (2) that the LHWCA did not apply because Virginia law could provide compensation. See id., at 245. As in Perini, our opinion reversing that decision did not discuss the Jones Act, because no one had even mentioned the possibility that the janitor might be a "seaman." Because Parker was decided during the 19-year period "during which the Court did not recognize the mutual exclusivity of the LHWCA and the Jones Act," Wilander, 498 U.S., at 348 ,7 it is not at all clear that the Court, if asked to do so, would not have found that the janitor was a Jones Act seaman as well as an LHWCA-covered employee. Accordingly, the cases cited by the majority lend no support to its holding that the member of a crew of a ship at sea is not always a seaman.

The Court's only other justification for refusing to apply a voyage test is its purported concern about a worker who might "walk into and out of coverage in the course of his regular duties." Ante at 15 (internal quotation marks omitted). Because the only way that a seaman could walk out of Jones Act coverage during a voyage would be to quit his job and become a passenger (or possibly jump overboard), I take the majority's argument to mean that a single voyage is not a long enough time to establish seaman status.8 I simply do not understand this argument. Surely a voyage is sufficient time to establish an employment-related, status-based connection to a vessel in navigation that exposes the employee to the perils of the sea. The majority cannot explain why an employee who signs on for a single journey is any less a "seaman" or "member of a crew" if he intends to become an insurance agent after the voyage than if he intends to remain with the ship. What is important is the employee's status at the time of the injury, not his status a day, a month, or a year beforehand or afterward.

Apparently, the majority's real concern about walking in and out of coverage is that an employer will be unable to predict which of his employees will be covered by the Jones Act, and which by the LHWCA, on any given day. I think it is a novel construction of the Jones Act to read it as a scheme to protect employers.9 But even if Congress had shared the Court's concern, this case does not implicate it in the least. We are talking here about a lengthy voyage on the high seas. The employer controls who goes on that voyage; he knows, more or less, when that voyage will begin and when it will end. And, but for the majority's decision today, he would know that while the ship is at sea, all his employees thereon would be covered by the Jones Act and not by the LHWCA. Thus, no one is walking out of Jones Act coverage and into LHWCA coverage (or vice versa) without the employer's knowledge and control. Once again, the majority's concern - and its method of determining seaman status - is properly directed at injuries occurring while the ship is at port.

As a matter of history, this concern with oscillating back and forth between different types of compensation systems recalls a very different and far more serious problem: the difficulty of defining who is a "maritime employee" (a class of workers that includes both seaman and longshoremen) and who is not. Over the powerful dissent of Justice Holmes, in Southern Pacific Co. v. Jensen, 244 U.S. 205 (1917), the Court held that the constitutional grant of admiralty and maritime jurisdiction to the federal courts prevented the State of New York from applying its workmen's compensation statute to a longshoreman who was injured on a gang plank about 10 feet seaward of Pier 49 in New York City. Jensen was a shore-based worker who had walked out of the coverage of the state law into an unprotected federal area - the area seaward of the shoreline. In enacting the LHWCA, Congress in 1927 responded to Jensen and its progeny by extending federal protection to shore-based workers injured while temporarily on navigable waters. The statute excluded Jones Act seamen, on the one hand, and shore-based workers while they were on the landward side of the Jensen line, on the other. As we have explained on more than one occasion, then, the LHWCA was originally a "gap-filling" measure intended to create coverage for those workers for whom, after Jensen, States could not provide compensation. See, e.g., Norton v. Warner Co., 321 U.S. 565, 570 (1944); Davis v. Department of Labor and Industries of Wash., 317 U.S. 249, 252 -253 (1942); see also S. Rep. No. 973, 69th Cong., 1st Sess., 16 (1926).10

Thus, the majority's concern about employees "walking in and out of coverage" evokes images of a real problem engendered by Jensen-the problem of employees changing their legal status, sometimes many times a day, merely by walking from one place to another in the course of their employment. That problem is not implicated in this case. At the time of his injury Latsis was employed, with the full knowledge of his employer, on a ship at sea. He could not walk out of coverage until the voyage was over. At the end of the voyage, if Latsis had taken on other duties, wholly or partly on land, and had been injured while so engaged, then the majority's concern might have substance. But in this case, the majority's concern - and its test for seaman status - is completely misplaced.
 

Footnotes

1. The "Jones Act" is actually 33 of the Merchant Marine Act, 1920, 41 Stat. 1007.

2. "Latsis's employment did expose him to the perils of the sea - in fact, his injury was the result of such a peril in the sense that while on board a seaman is very much reliant upon and in the care of the ship's physician. If that physician is unqualified or engages in medical malpractice, it is just as much a peril to the mariner on board as the killer wave, the gale or hurricane, or other dangers of the calling." 20 F.3d 45, 55 (CA2 1994).

3. These remedies are maintenance and cure and recovery for unseaworthiness. See G. Gilmore & C. Black, Law of Admiralty, ch. VI (2d ed. 1975).

4. In Desper, we held that a workman on a moored barge was not a "seaman" at the time of his death even though "he was a probable navigator in the near future." 342 U.S., at 191 . We noted that "[t]he many cases turning upon the question whether an individual was a `seaman' demonstrate that the matter depends largely on the facts of the particular case and the activity in which he was engaged at the time of injury. . . . [T]here was no vessel engaged in navigation at the time of the decedent's death." Id., at 190-191.

5. The quotation is from a pre-Jones Act case, The Buena Ventura, 243 F. 797, 799 (SDNY 1916). Earlier in his opinion, Justice Cardozo had noted: "In the enforcement of the statute a policy of liberal construction announced at the beginning has been steadily maintained." Warner, 293 U.S., at 156.

6. Even in Warner, no one contested the basic proposition that an employee of a ship at sea is a "seaman." Instead, the issue in that case was whether the term "seaman" extended to the captain of such a ship, or whether it referred only to lower-level employees. The Court, applying the "liberal construction" that Congress intended, held that the master was a "seaman."

7. During this period, the Court incorrectly treated stevedores working on moored vessels as seamen covered by the Jones Act under the pre-LHWCA ruling in International Stevedoring Co. v. Haverty, 272 U.S. 50 (1926). See Wilander, 498 U.S., at 348 -349.

8. Or at least, it is not necessarily a long enough time. It depends on the facts. See ante, at 24-25.

9. The Jones Act was passed to overturn the harsh rule of The Osceola, 189 U.S. 158 (1903), which disallowed any recovery by a seaman for negligence of the master or any member of the crew of his ship under general maritime law. McDermott International, Inc. v. Wilander, 498 U.S. 337 , at 342 (1991). The aim of the statute, then, was to expand the remedies available to employees, not to aid their employers.

10. Whereas the LHWCA as enacted in 1927 responded to the problem of employees who walked out of state coverage every time they boarded a ship, the 1972 Amendment to that Act responded to the opposite concern longshoremen who walked out of federal coverage every time they left the ship. Because state compensation schemes were sometimes less generous than the LHWCA, Congress expanded the federal coverage to encompass injuries occurring on piers and adjacent land used for loading and unloading ships. See H. R. Rep. No. 92-1441, pp. 10-11 (1972). Because the class of workers protected by the LHWCA continued to be composed entirely of shore-based workers, the 1972 amendment appropriately preserved the exclusion of Jones Act seamen. It did not alter the original 1927 Act's constructive definition of "seaman" as "master or member of a crew of any vessel."

11.  R. Stevenson, Treasure Island (1883).

12.   H. Melville, Billy Budd (1924).