delivered the opinion of the Court.

At issue in this case is the validity of an exclusive contract between a hospital and a firm of anesthesiologists. We must decide whether the contract gives rise to a per se violation of § 1 of the Sherman Act1 because every patient undergoing *5surgery at the hospital must use the services of one firm of anesthesiologists, and, if not, whether the contract is nevertheless illegal because it unreasonably restrains competition among anesthesiologists.

In July 1977, respondent Edwin G. Hyde, a board-certified anesthesiologist, applied for admission to the medical staff of East Jefferson Hospital. The credentials committee and the medical staff executive committee recommended approval, but the hospital board denied the application because the hospital was a party to a contract providing that all anesthesio-logical services required by the hospital’s patients would be performed by Roux & Associates, a professional medical corporation. Respondent then commenced this action seeking a declaratory judgment that the contract is unlawful and an injunction ordering petitioners to appoint him to the hospital staff.2 After trial, the District Court denied relief, finding that the anticompetitive consequences of the Roux contract were minimal and outweighed by benefits in the form of improved patient care. 513 F. Supp. 532 (ED La. 1981). The Court of Appeals reversed because it was persuaded that the contract was illegal “per se. ” 686 F. 2d 286 (CA5 1982). We granted certiorari, 460 U. S. 1021 (1983), and now reverse.

I

In February 1971, shortly before East Jefferson Hospital opened, it entered into an “Anesthesiology Agreement” with Roux & Associates (Roux), a firm that had recently been organized by Dr. Kermit Roux. The contract provided that any anesthesiologist designated by Roux would be admitted to the hospital’s medical staff. The hospital agreed to *6provide the space, equipment, maintenance, and other supporting services necessary to operate the anesthesiology department. It also agreed to purchase all necessary drugs and other supplies. All nursing personnel required by the anesthesia department were to be supplied by the hospital, but Roux had the right to approve their selection and retention.3 The hospital agreed to “restrict the use of its anesthesia department to Roux & Associates and [that] no other persons, parties or entities shall perform such services within the Hospital for the ter[m] of this contract.” App. 19.4

The 1971 contract provided for a 1-year term automatically renewable for successive 1-year periods unless either party elected to terminate. In 1976, a second written contract was executed containing most of the provisions of the 1971 agreement. Its term was five years and the clause excluding other anesthesiologists from the hospital was deleted;5 the hospital nevertheless continued to regard itself as committed to a closed anesthesiology department. Only Roux was permitted to practice anesthesiology at the hospital. At the *7time of trial the department included four anesthesiologists. The hospital usually employed 13 or 14 certified registered nurse anesthetists.6

The exclusive contract had an impact on two different segments of the economy: consumers of medical services, and providers of anesthesiological services. Any consumer of medical services who elects to have an operation performed at East Jefferson Hospital may not employ any anesthesiologist not associated with Roux. No anesthesiologists except those employed by Roux may practice at East Jefferson.

There are at least 20 hospitals in the New Orleans metropolitan area and about 70 percent of the patients living in Jefferson Parish go to hospitals other than East Jefferson. Because it regarded the entire New Orleans metropolitan area as the relevant geographic market in which hospitals compete, this evidence convinced the District Court that East Jefferson does not possess any significant “market power”; therefore it concluded that petitioners could not use the Roux contract to anticompetitive ends.7 The same evidence led the Court of Appeals to draw a different conclusion. Noting that 30 percent of the residents of the parish go to East Jefferson Hospital, and that in fact “patients tend to choose hospitals by location rather than price or quality,” the Court of *8Appeals concluded that the relevant geographic market was the East Bank of Jefferson Parish. 686 F. 2d, at 290. The conclusion that East Jefferson Hospital possessed market power in that area was buttressed by the facts that the prevalence of health insurance eliminates a patient’s incentive to compare costs, that the patient is not sufficiently informed to compare quality, and that family convenience tends to magnify the importance of location.8

The Court of Appeals held that the case involves a “tying arrangement” because the “users of the hospital’s operating rooms (the tying product) are also compelled to purchase the hospital’s chosen anesthesia service (the tied product).” Id., at 289. Having defined the relevant geographic market for the tying product as the East Bank of Jefferson Parish, the court held that the hospital possessed “sufficient market power in the tying market to coerce purchasers of the tied product.” Id., at 291. Since the purchase of the tied product constituted a “not insubstantial amount of interstate commerce,” under the Court of Appeals’ reading of our decision in Northern Pacific R. Co. v. United States, 356 U. S. 1, 11 (1958), the tying arrangement was therefore illegal “per se.”9

*9i — H 1 — 4

Certain types of contractual arrangements are deemed unreasonable as a matter of law.10 The character of the restraint produced by such an arrangement is considered a sufficient basis for presuming unreasonableness without the necessity of any analysis of the market context in which the arrangement may be found.11 A price-fixing agreement between competitors is the classic example of such an arrangement. Arizona v. Maricopa County Medical Society, 457 U. S. 332, 343-348 (1982). It is far too late in the history of our antitrust jurisprudence to question the proposition that certain tying arrangements pose an unacceptable risk of stifling competition and therefore are unreasonable “per se.”12 The rule was first enunciated in International Salt Co. v. United States, 332 U. S. 392, 396 (1947),13 and has been en*10dorsed by this Court many times since.14 The rule also reflects congressional policies underlying the antitrust laws. In enacting § 3 of the Clayton Act, 38 Stat. 731, 15 U. S. C. § 14, Congress expressed great concern about the anti-competitive character of tying arrangements. See H. R. Rep. No. 627, 63d Cong., 2d Sess., 10-13 (1914); S. Rep. No. 698, 63d Cong., 2d Sess., 6-9 (1914).15 While this case *11does not arise under the Clayton Act, the congressional finding made therein concerning the competitive consequences of tying is illuminating, and must be respected.16

It is clear, however, that not every refusal to sell two products separately can be said to restrain competition. If each of the products may be purchased separately in a competitive market, one seller’s decision to sell the two in a single package imposes no unreasonable restraint on either market, par*12ticularly if competing suppliers are free to sell either the entire package or its several parts.17 For example, we have written that “if one of a dozen food stores in a community were to refuse to sell flour unless the buyer also took sugar it would hardly tend to restrain competition in sugar if its competitors were ready and able to sell flour by itself.” Northern Pacific R. Co. v. United States, 356 U. S., at 7.18 Buyers often find package sales attractive; a seller’s decision to offer such packages can merely be an attempt to compete effectively — conduct that is entirely consistent with the Sherman Act. See Fortner Enterprises v. United States Steel Corp., 394 U. S. 495, 517-518 (1969) (Fortner I) (White, J., dissenting); id., at 524-525 (Fortas, J., dissenting).

Our cases have concluded that the essential characteristic of an invalid tying arrangement lies in the seller’s exploitation of its control over the tying product to force the buyer into the purchase of a tied product that the buyer either did not want at all, or might have preferred to purchase elsewhere on different terms. When such “forcing” is present, competition on the merits in the market for the tied item is restrained and the Sherman Act is violated.

“Basic to the faith that a free economy best promotes the public weal is that goods must stand the cold test of competition; that the public, acting through the market’s impersonal judgment, shall allocate the Nation’s resources and thus direct the course its economic development will take. ... By conditioning his sale of one commodity on *13the purchase of another, a seller coerces the abdication of buyers’ independent judgment as to the ‘tied’ product’s merits and insulates it from the competitive stresses of the open market. But any intrinsic superiority of the ‘tied’ product would convince freely choosing buyers to select it over others anyway.” Times-Picayune Publishing Co. v. United States, 345 U. S. 594, 605 (1953).19

Accordingly, we have condemned tying arrangements when the seller has some special ability — usually called “mar*14ket power” — to force a purchaser to do something that he would not do in a competitive market. See United States Steel Corp. v. Fortner Enterprises, 429 U. S. 610, 620 (1977) (Fortner II); Fortner I, 394 U. S., at 503-504; United States v. Loew’s Inc., 371 U. S. 38, 45, 48, n. 5 (1962); Northern Pacific R. Co. v. United States, 356 U. S., at 6-7.20 When “forcing” occurs, our cases have found the tying arrangement to be unlawful.

Thus, the law draws a distinction between the exploitation of market power by merely enhancing the price of the tying product, on the one hand, and by attempting to impose restraints on competition in the market for a tied product, on the other. When the seller’s power is just used to maximize its return in the tying product market, where presumably its product enjoys some justifiable advantage over its competitors, the competitive ideal of the Sherman Act is not necessarily compromised. But if that power is used to impair competition on the merits in another market, a potentially inferior product may be insulated from competitive pressures.21 This impairment could either harm existing competitors or create barriers to entry of new competitors in the market for the tied product, Fortner I, 394 U. S., at 509,22 and can in*15crease the social costs of market power by facilitating price discrimination, thereby increasing monopoly profits over what they would be absent the tie, Fortner II, 429 U. S., at 617.23 And from the standpoint of the consumer — whose interests the statute was especially intended to serve — the freedom to select the best bargain in the second market is impaired by his need to purchase the tying product, and perhaps by an inability to evaluate the true cost of either product when they are available only as a package.24 In sum, to permit restraint of competition on the merits through tying arrangements would be, as we observed in Fortner II, to condone “the existence of power that a free market would not tolerate.” 429 U. S., at 617 (footnote omitted).

Per se condemnation — condemnation without inquiry into actual market conditions — is only appropriate if the existence of forcing is probable.25 Thus, application of the per se rule *16focuses on the probability of anticompetitive consequences. Of course, as a threshold matter there must be a substantial potential for impact on competition in order to justify per se condemnation. If only a single purchaser were “forced” with respect to the purchase of a tied item, the resultant impact on competition would not be sufficient to warrant the concern of antitrust law. It is for this reason that we have refused to condemn tying arrangements unless a substantial volume of commerce is foreclosed thereby. See Fortner I, 394 U. S., at 501-502; Northern Pacific R. Co. v. United States, 356 U. S., at 6-7; Times-Picayune, 345 U. S., at 608-610; International Salt, 332 U. S., at 396. Similarly, when a purchaser is “forced” to buy a product he would not have otherwise bought even from another seller in the tied-product market, there can be no adverse impact on competition because no portion of the market which would otherwise have been available to other sellers has been foreclosed.

Once this threshold is surmounted, per se prohibition is appropriate if anticompetitive forcing is likely. For example, if the Government has granted the seller a patent or similar monopoly over a product, it is fair to presume that the inability to buy the product elsewhere gives the seller market power. United States v. Loew’s Inc., 371 U. S., at 45-47. Any effort to enlarge the scope of the patent monopoly by using the market power it confers to restrain competition in the market for a second product will undermine competition on the merits in that second market. Thus, the sale or lease of a patented item on condition that the buyer make all his purchases of a separate tied product from the patentee is unlawful. See United States v. Paramount Pictures, Inc., 334 U. S. 131, 156-159 (1948); International Salt, 332 *17U. S., at 395-396; International Business Machines Corp. v. United States, 298 U. S. 131 (1936).

The same strict rule is appropriate in other situations in which the existence of market power is probable. When the seller’s share of the market is high, see Times-Picayune Publishing Co. v. United States, 345 U. S., at 611-613, or when the seller offers a unique product that competitors are not able to offer, see Fortner I, 394 U. S., at 504-506, and n. 2, the Court has held that the likelihood that market power exists and is being used to restrain competition in a separate market is sufficient to make per se condemnation appropriate. Thus, in Northern Pacific R. Co. v. United States, 356 U. S. 1 (1958), we held that the railroad’s control over vast tracts of western real estate, although not itself unlawful, gave the railroad a unique kind of bargaining power that enabled it to tie the sales of that land to exclusive, long-term commitments that fenced out competition in the transportation market over a protracted period.26 When, however, the *18seller does not have either the degree or the kind of market power that enables him to force customers to purchase a second, unwanted product in order to obtain the tying product, an antitrust violation can be established only by evidence of an unreasonable restraint on competition in the relevant market. See Fortner I, 394 U. S., at 499-500; Times-Picayune Publishing Co. v. United States, 345 U. S., at 614-615.

In sum, any inquiry into the validity of a tying arrangement must focus on the market or markets in which the two products are sold, for that is where the anticompetitive forcing has its impact. Thus, in this case our analysis of the tying issue must focus on the hospital’s sale of services to its patients, rather than its contractual arrangements with the providers of anesthesiological services. In making that analysis, we must consider whether petitioners are selling two separate products that may be tied together, and, if so, whether they have used their market power to force their patients to accept the tying arrangement.

III

The hospital has provided its patients with a package that includes the range of facilities and services required for a variety of surgical operations.27 At East Jefferson Hospital the package includes the services of the anesthesiologist.28 Petitioners argue that the package does not involve a tying ar*19rangement at all — that they are merely providing a functionally integrated package of services.29 Therefore, petitioners contend that it is inappropriate to apply principles concerning tying arrangements to this case.

Our cases indicate, however, that the answer to the question whether one or two products are involved turns not on the functional relation between them, but rather on the character of the demand for the two items.30 In Times-Picayune Publishing Co. v. United States, 345 U. S. 594 (1953), the Court held that a tying arrangement was not present because the arrangement did not link two distinct markets for products that were distinguishable in the eyes of buyers.31 In *20Fortner I, the Court concluded that a sale involving two independent transactions, separately priced and purchased from the buyer’s perspective, was a tying arrangement.32 These *21cases make it clear that a tying arrangement cannot exist unless two separate product markets have been linked.

The requirement that two distinguishable product markets be involved follows from the underlying rationale of the rule against tying. The definitional question depends on whether the arrangement may have the type of competitive consequences addressed by the rule.33 The answer to the question whether petitioners have utilized a tying arrangement must be based on whether there is a possibility that the economic effect of the arrangement is that condemned by the rule against tying — that petitioners have foreclosed competition on the merits in a product market distinct from the market for the tying item.34 Thus, in this case no tying arrangement can exist unless there is a sufficient demand for the purchase of anesthesiological services separate from hospital services *22to identify a distinct product market in which it is efficient to offer anesthesiological services separately from hospital services.35

Unquestionably, the anesthesiological component of the package offered by the hospital could be provided separately and could be selected either by the individual patient or by one of the patient’s doctors if the hospital did not insist on including anesthesiological services in the package it offers to its customers. As a matter of actual practice, anesthesiological services are billed separately from the hospital services petitioners provide. There was ample and uncontroverted testimony that patients or surgeons often request specific anesthesiologists to come to a hospital and provide anesthesia, and that the choice of an individual anesthesiologist separate from the choice of a hospital is particularly frequent in respondent’s specialty, obstetric anesthesiology.36 The Dis*23trict Court found that “[t]he provision of anesthesia services is a medical service separate from the other services provided by the hospital.” 513 F. Supp., at 540.37 The Court of Appeals agreed with this finding, and went on to observe: “[A]n anesthesiologist is normally selected by the surgeon, rather than the patient, based on familiarity gained through a working relationship. Obviously, the surgeons who practice at East Jefferson Hospital do not gain familiarity with any anesthesiologists other than Roux and Associates.” 686 F. 2d, at 291.38 The record amply supports the conclusion that consumers differentiate between anesthesiological services and the other hospital services provided by petitioners.39

*24Thus, the hospital’s requirement that its patients obtain necessary anesthesiological services from Roux combined the purchase of two distinguishable services in a single transaction.40 Nevertheless, the fact that this case involves a re*25quired purchase of two services that would otherwise be purchased separately does not make the Roux contract illegal. As noted above, there is nothing inherently anticompetitive about packaged sales. Only if patients are forced to purchase Roux’s services as a result of the hospital’s market power would the arrangement have anticompetitive consequences. If no forcing is present, patients are free to enter a competing hospital and to use another anesthesiologist instead of Roux.41 The fact that petitioners’ patients are required to purchase two separate items is only the beginning of the appropriate inquiry.42

*26I — < <

The question remains whether this arrangement involves the use of market power to force patients to buy services they would not otherwise purchase. Respondent’s only basis for invoking the per se rule against tying and thereby avoiding analysis of actual market conditions is by relying on the preference of persons residing in Jefferson Parish to go to East Jefferson, the closest hospital. A preference of this kind, however, is not necessarily probative of significant market power.

Seventy percent of the patients residing in Jefferson Parish enter hospitals other than East Jefferson. 513 F. Supp., at 539. Thus East Jefferson’s “dominance” over persons residing in Jefferson Parish is far from overwhelming.43 The *27fact that a substantial majority of the parish’s residents elect not to enter East Jefferson means that the geographic data do not establish the kind of dominant market position that obviates the need for further inquiry into actual competitive conditions. The Court of Appeals acknowledged as much; it recognized that East Jefferson’s market share alone was insufficient as a basis to infer market power, and buttressed its conclusion by relying on “market imperfections”44 that permit petitioners to charge noncompetitive prices for hospital services: the prevalence of third-party payment for health care costs reduces price competition, and a lack of adequate information renders consumers unable to evaluate the quality of the medical care provided by competing hospitals. 686 F. 2d, at 290.45 While these factors may generate “market power” in some abstract sense,46 they do not generáte the kind of market power that justifies condemnation of tying.

Tying arrangements need only be condemned if they restrain competition on the merits by forcing purchases that would not otherwise be made. A lack of price or quality *28competition does not create this type of forcing. If consumers lack price consciousness, that fact will not force them to take an anesthesiologist whose services they do not want— their indifference to price will have no impact on their willingness or ability to go to another hospital where they can utilize the services of the anesthesiologist of their choice. Similarly, if consumers cannot evaluate the quality of an-esthesiological services, it follows that they are indifferent between certified anesthesiologists even in the absence of a tying arrangement — such an arrangement cannot be said to have foreclosed a choice that would have otherwise been made “on the merits.”

Thus, neither of the “market imperfections” relied upon by the Court of Appeals forces consumers to take anesthesiological services they would not select in the absence of a tie. It is safe to assume that every patient undergoing a surgical operation needs the services of an anesthesiologist; at least this record contains no evidence that the hospital “forced” any such services on unwilling patients.47 The record therefore *29does not provide a basis for applying the per se rule against tying to this arrangement.

V

In order to prevail in the absence of per se liability, respondent has the burden of proving that the Roux contract violated the Sherman Act because it unreasonably restrained competition. That burden necessarily involves an inquiry into the actual effect of the exclusive contract on competition among anesthesiologists. This competition takes place in a market that has not been defined. The market is not necessarily the same as the market in which hospitals compete in offering services to patients; it may encompass competition among anesthesiologists for exclusive contracts such as the Roux contract and might be statewide or merely local.48 There is, however, insufficient evidence in this record to provide a basis for finding that the Roux contract, as it actually operates in the market, has unreasonably restrained compe*30tition. The record sheds little light on how this arrangement affected consumer demand for separate arrangements with a specific anesthesiologist.49 The evidence indicates that some surgeons and patients preferred respondent’s services to those of Roux, but there is no evidence that any patient who was sophisticated enough to know the difference between two anesthesiologists was not also able to go to a hospital that would provide him with the anesthesiologist of his choice.50

In sum, all that the record establishes is that the choice of anesthesiologists at East Jefferson has been limited to one of the four doctors who are associated with Roux and therefore have staff privileges.51 Even if Roux did not have an exclusive contract, the range of alternatives open to the patient would be severely limited by the nature of the transaction and the hospital’s unquestioned right to exercise some control over the identity and the number of doctors to whom it accords staff privileges. If respondent is admitted to the staff of East Jefferson, the range of choice will be enlarged from *31four to five doctors, but the most significant restraints on the patient’s freedom to select a specific anesthesiologist will nevertheless remain.52 Without a showing of actual adverse effect on competition, respondent cannot make out a case under the antitrust laws, and no such showing has been made.

VI

Petitioners’ closed policy may raise questions of medical ethics,53 and may have inconvenienced some patients who would prefer to have their anesthesia administered by someone other than a member of Roux & Associates, but it does not have the obviously unreasonable impact on purchasers that has characterized the tying arrangements that this Court has branded unlawful. There is no evidence that the price, the quality, or the supply or demand for either the “tying product” or the “tied product” involved in this case has been adversely affected by the exclusive contract between Roux and the hospital. It may well be true that the contract made it necessary for Dr. Hyde and others to practice elsewhere, rather than at East Jefferson. But there has been no showing that the market as a whole has been affected at all by the contract. Indeed, as we previously noted, the record tells us very little about the market for the services of an*32esthesiologists. Yet that is the market in which the exclusive contract has had its principal impact. There is simply no showing here of the kind of restraint on competition that is prohibited by the Sherman Act. Accordingly, the judgment of the Court of Appeals is reversed, and the case is remanded to that court for further proceedings consistent with this opinion.54

It is so ordered.

with whom Justice Marshall joins, concurring.

As the opinion for the Court demonstrates, we have long held that tying arrangements are subject to evaluation for per se illegality under § 1 of the Sherman Act. Whatever merit the policy arguments against this longstanding construction of the Act might have, Congress, presumably aware of our decisions, has never changed the rule by amending the Act. In such circumstances, our practice usually has been to stand by a settled statutory interpretation and leave the task of modifying the statute’s reach to Congress. See Monsanto Co. v. Spray-Rite Service Corp., 465 U. S. 752, 769 (1984) (Brennan, J., concurring). I see no reason to depart from that principle in this case and therefore join the opinion and judgment of the Court.

with whom The Chief Justice, Justice Powell, and Justice Rehnquist join, concurring in the judgment.

East Jefferson Hospital, a public hospital governed by petitioners, requires patients to use the anesthesiological services provided by Roux & Associates, as they are the only doctors authorized to administer anesthesia to patients in the hospital. The Court of Appeals found that this arrangement was a tie-in illegal under the Sherman Act. 686 F. 2d 286 *33(CA5 1982). I concur in the Court’s decision to reverse but write separately to explain why I believe the hospital-Roux contract, whether treated as effecting a tie between services provided to patients, or as an exclusive dealing arrangement between the hospital and certain anesthesiologists, is properly analyzed under the rule of reason.

hH

Tying is a form of marketing in which a seller insists on selling two distinct products or services as a package. A supermarket that will sell flour to consumers only if they will also buy sugar is engaged in tying. Flour is referred to as the tying product, sugar as the tied product. In this case the allegation is that East Jefferson Hospital has unlawfully tied the sale of general hospital services and operating room facilities (the tying service) to the sale of anesthesiologists’ services (the tied services). The Court has on occasion applied a per se rule of illegality in actions alleging tying in violation of § 1 of the Sherman Act. International Salt Co. v. United States, 332 U. S. 392 (1947).

Under the usual logic of the per se rule, a restraint on trade that rarely serves any purposes other than to restrain competition is illegal without proof of market power or anticompet-itive effect. See, e. g., Northern Pacific R. Co. v. United States, 356 U. S. 1, 5 (1958). In deciding whether an economic restraint should be declared illegal per se, “[t]he probability that anticompetitive consequences will result from a practice and the severity of those consequences [is] balanced against its procompetitive consequences. Cases that do not fit the generalization may arise, but a per se rule reflects the judgment that such cases are not sufficiently common or important to justify the time and expense necessary to identify them.” Continental T. V., Inc. v. GTE Sylvania Inc., 433 U. S. 36, 50, n. 16 (1977). See also Arizona v. Maricopa County Medical Society, 457 U. S. 332, 351 (1982). Only when there is very little loss to society from banning a re*34straint altogether is an inquiry into its costs in the individual case considered to be unnecessary.

Some of our earlier cases did indeed declare that tying arrangements serve “hardly any purpose beyond the suppression of competition.” Standard Oil Co. of California v. United States, 337 U. S. 293, 305-306 (1949) (dictum). However, this declaration was not taken literally even by the cases that purported to rely upon it. In practice, a tie has been illegal only if the seller is shown to have “sufficient economic power with respect to the tying product to appreciably restrain free competition in the market for the tied product . . . .” Northern Pacific R. Co., 356 U. S., at 6. Without “control or dominance over the tying product,” the seller could not use the tying product as “an effectual weapon to pressure buyers into taking the tied item,” so that any restraint of trade would be “insignificant.” Ibid. The Court has never been willing to say of tying arrangements, as it has of price fixing, division of markets, and other agreements subject to per se analysis, that they are always illegal, without proof of market power or anticompetitive effect.

The “per se” doctrine in tying cases has thus always required an elaborate inquiry into the economic effects of the tying arrangement.1 As a result, tying doctrine incurs the costs of a rule-of-reason approach without achieving its benefits: the doctrine calls for the extensive and time-consuming economic analysis characteristic of the rule of reason, but then may be interpreted to prohibit arrangements that economic analysis would show to be beneficial. Moreover, the per se label in the tying context has generated more confusion *35than coherent law because it appears to invite lower courts to omit the analysis of economic circumstances of the tie that has always been a necessary element of tying analysis.

The time has therefore come to abandon the “per se” label and refocus the inquiry on the adverse economic effects, and the potential economic benefits, that the tie may have. The law of tie-ins will thus be brought into accord with the law applicable to all other allegedly anticompetitive economic arrangements, except those few horizontal or quasi-horizontal restraints that can be said to have no economic justification whatsoever.2 This change will rationalize rather than abandon tie-in doctrine as it is already applied.

I — I » — I

Our prior opinions indicate that the purpose of tying law has been to identify and control those tie-ins that have a demonstrable exclusionary impact in the tied-product market, see Times-Picayune Publishing Co. v. United States, 345 U. S. 594, 605 (1953), or that abet the harmful exercise of market power that the seller possesses in the tying product market.3 Under the rule of reason tying arrangements should be disapproved only in such instances.

Market power in the tying product may be acquired legitimately (e. g., through the grant of a patent) or illegitimately (e. g., as a result of unlawful monopolization). In either event, exploitation of consumers in the market for the tying *36product is a possibility that exists and that may be regulated under § 2 of the Sherman Act without reference to any tying arrangements that the seller may have developed. The existence of a tied product normally does not increase the profit that the seller with market power can extract from sales of the tying product. A seller with a monopoly on flour, for example, cannot increase the profit it can extract from flour consumers simply by forcing them to buy sugar along -with their flour. Counterintuitive though that assertion may seem, it is easily demonstrated and widely accepted. See, e. g., R. Bork, The Antitrust Paradox 372-374 (1978); P. Areeda, Antitrust Analysis 735 (3d ed. 1981).

Tying may be economically harmful primarily in the rare cases where power in the market for the tying product is used to create additional market power in the market for the tied product.4 The antitrust law is properly concerned -with *37tying when, for example, the flour monopolist threatens to use its market power to acquire additional power in the sugar market, perhaps by driving out competing sellers of sugar, or by making it more difficult for new sellers to enter the sugar market. But such extension of market power is unlikely, or poses no threat of economic harm, unless the two markets in question and the nature of the two products tied satisfy three threshold criteria.5

First, the seller must have power in the tying-product market.6 Absent such power tying cannot conceivably have any adverse impact in the tied-product market, and can be only procompetitive in the tying-product market.7 If the *38seller of flour has no market power over flour, it will gain none by insisting that its buyers take some sugar as well. See United States Steel Corp. v. Fortner Enterprises, Inc., 429 U. S. 610, 620 (1977) (Fortner II); Fortner Enterprises, Inc. v. United States Steel Corp., 394 U. S. 496, 503-504 (1969) (Fortner I); United States v. Loew’s Inc., 371 U. S. 38, 45, 48, n. 5 (1962); Northern Pacific R. Co. v. United States, 356 U. S., at 6-7.

Second, there must be a substantial threat that the tying seller will acquire market power in the tied-product market. No such threat exists if the tied-product market is occupied by many stable sellers who are not likely to be driven out by the tying, or if entry barriers in the tied-product market are low. If, for example, there is an active and vibrant market for sugar — one with numerous sellers and buyers who do not deal in flour — the flour monopolist’s tying of sugar to flour need not be declared unlawful. Cf. Fortner II, supra, at 617-618, and n. 8; Fortner I, supra, at 498-499; Times-Picayune Publishing Co. v. United States, 345 U. S., at 611; Standard Oil Co. of California v. United States, 337 U. S., at 305-306; International Salt Co. v. United States, 332 *39U. S., at 396. If, on the other hand, the tying arrangement is likely to erect significant barriers to entry into the tied-product market, the tie remains suspect. Atlantic Refining Co. v. FTC, 381 U. S. 357, 371 (1965).

Third, there must be a coherent economic basis for treating the tying and tied products as distinct. All but the simplest products can be broken down into two or more components that are “tied together” in the final sale. Unless it is to be illegal to sell cars with engines or cameras with lenses, this analysis must be guided by some limiting principle. For products to be treated as distinct, the tied product must, at a minimum, be one that some consumers might wish to purchase separately without also purchasing the tying product.8 When the tied product has no use other than in conjunction with the tying product, a seller of the tying product can acquire no additional market power by selling the two products together. If sugar is useless to consumers except when used with flour, the flour seller’s market power is projected into the sugar market whether or not the two products are actually sold together; the flour seller can exploit what market power it has over flour with or without the tie.9 The flour seller will therefore have little incentive to monopolize the sugar market unless it can produce and distribute sugar more cheaply than other sugar sellers. And in this unusual case, where flour is monopolized and sugar is useful only when *40used with flour, consumers -will suffer no further economic injury by the monopolization of the sugar market.

Even when the tied product does have a use separate from the tying product, it makes little sense to label a package as two products without also considering the economic justifications for the sale of the package as a unit. When the economic advantages of joint packaging are substantial the package is not appropriately viewed as two products, and that should be the end of the tying inquiry. The lower courts largely have adopted this approach.10 See, e. g., Foster v. Maryland State Savings and Loan Assn., 191 U. S. App. D. C. 226, 228-231, 590 F. 2d 928, 930-933 (1978), cert. denied, 439 U. S. 1071 (1979); Response of Carolina, Inc. v. Leasco Response, Inc., 537 F. 2d 1307, 1330 (CA5 1976); Kugler v. AAMCO Automatic Transmissions, Inc., 460 F. 2d 1214 (CA8 1972); ILC Peripherals Leasing Corp. v. International Business Machines Corp., 448 F. Supp. 228, 230 *41(ND Cal. 1978); United States v. Jerrold Electronics Corp., 187 F. Supp. 545, 563 (ED Pa. 1960), aff’d per curiam, 365 U. S. 567 (1961).

These three conditions — market power in the tying product, a substantial threat of market power in the tied product, and a coherent economic basis for treating the products as distinct — are only threshold requirements. Under the rule of reason a tie-in may prove acceptable even when all three are met. Tie-ins may entail economic benefits as well as economic harms, and if the threshold requirements are met these benefits should enter the rule-of-reason balance.

“[Tie-ins] may facilitate new entry into fields where established sellers have wedded their customers to them by ties of habit and custom. Brown Shoe Co. v. United States, 370 U. S. 294, 330 (1962) .... They may permit clandestine price cutting in products which otherwise would have no price competition at all because of fear of retaliation from the few other producers dealing in the market. They may protect the reputation of the tying product if failure to use the tied product in conjunction with it may cause it to misfunction. . . . [Citing] Pick Mfg. Co. v. General Motors Corp., 80 F. 2d 641 (C. A. 7th Cir. 1935), aff’d, 299 U. S. 3 (1936). And, if the tied and tying products are functionally related, they may reduce costs through economies of joint production and distribution.” Fortner I, 394 U. S., at 514, n. 9 (White, J., dissenting).

The ultimate decision whether a tie-in is illegal under the antitrust laws should depend upon the demonstrated economic effects of the challenged agreement. It may, for example, be entirely innocuous that the seller exploits its control over the tying product to “force” the buyer to purchase the tied product. For when the seller exerts market power only in the tying-product market, it makes no difference to him or his customers whether he exploits that power by rais*42ing the price of the tying product or by “forcing” customers to buy a tied product. See Markovits, Tie-Ins, Reciprocity and the Leverage Theory, 76 Yale L. J. 1397, 1397-1398 (1967); Burstein, A Theory of Full-Line Forcing, 55 Nw. U. L. Rev. 62, 62-63 (1960). On the other hand, tying may make the provision of packages of goods and services more efficient. A tie-in should be condemned only when its anticompetitive impact outweighs its contribution to efficiency.

III

Application of these criteria to the case at hand is straightforward.

Although the issue is in doubt, we may assume that the hospital does have market power in the provision of hospital services in its area. The District Court found to the contrary, 513 F. Supp. 532, 541 (ED La. 1981), but the Court of Appeals determined that the hospital does possess market power in an appropriately defined market. While appellate courts should normally defer to the district courts’ findings on such fact-bound questions,11 I shall assume for the purposes of this discussion that the Court of Appeals' determination that the hospital does have some power in the provision of hospital services in its local market is accepted.

Second, in light of the hospital’s presumed market power, we may also assume that there is a substantial threat that East Jefferson will acquire market power over the provision of anesthesiological services in its market. By tying the sale of anesthesia to the sale of other hospital services the hospital can drive out other sellers of those services who might otherwise operate in the local market. The hospital may thus gain local market power in the provision of anesthesiology: an-esthesiological services offered in the hospital’s market, narrowly defined, will be purchased only from Roux, under the hospital’s auspices.

*43But the third threshold condition for giving closer scrutiny to a tying arrangement is not satisfied here: there is no sound economic reason for treating surgery and anesthesia as separate services. Patients are interested in purchasing anesthesia only in conjunction with hospital services,12 so the hospital can acquire no additional market power by selling the two services together. Accordingly, the link between the hospital’s services and anesthesia administered by Roux will affect neither the amount of anesthesia provided nor the combined price of anesthesia and surgery for those who choose to become the hospital’s patients. In these circumstances, anesthesia and surgical services should probably not be characterized as distinct products for tying purposes.

Even if they are, the tying should not be considered a violation of § 1 of the Sherman Act because tying here cannot increase the seller’s already absolute power over the volume of production of the tied product, which is an inevitable consequence of the fact that very few patients will choose to undergo surgery without receiving anesthesia. The hospital-Roux contract therefore has little potential to harm the patients. On the other side of the balance, the District Court found, and the Court of Appeals did not dispute, that the tie-in conferred significant benefits upon the hospital and the patients that it served.

The tie-in improves patient care and permits more efficient hospital operation in a number of ways. From the viewpoint of hospital management, the tie-in ensures 24-hour anesthesiology coverage, aids in standardization of procedures and efficient use of equipment, facilitates flexible scheduling of operations, and permits the hospital more effectively to monitor the quality of anesthesiological services. Further, the tying arrangement is advantageous to patients because, as the District Court found, the closed anesthesiology depart*44ment places upon the hospital, rather than the individual patient, responsibility to select the physician who is to provide anesthesiological services. The hospital also assumes the responsibility that the anesthesiologist will be available, will be acceptable to the surgeon, and will provide suitable care to the patient. In assuming these responsibilities — responsibilities that a seriously ill patient frequently may be unable to discharge — the hospital provides a valuable service to its patients. And there is no indication that patients were dissatisfied with the quality of anesthesiology that was provided at the hospital or that patients wished to enjoy the services of anesthesiologists other than those that the hospital employed. Given this evidence of the advantages and effectiveness of the closed anesthesiology department, it is not surprising that, as the District Court found, such arrangements are accepted practice in the majority of hospitals of New Orleans and in the health care industry generally. Such an arrangement, which has little anticompetitive effect and achieves substantial benefits in the provision of care to patients, is hardly one that the antitrust law should condemn.13 This conclusion reaffirms our threshold determination that the joint provision of hospital services and anesthesiology should not be viewed as involving a tie between distinct products, and therefore should require no additional scrutiny under the antitrust law.

> hH

Whether or not the hospital-Roux contract is characterized as a tie between distinct products, the contract unquestionably does constitute exclusive dealing. Exclusive-dealing arrangements are independently subject to scrutiny under § 1 of the Sherman Act, and are also analyzed under the rule of *45reason. Tampa Electric Co. v. Nashville Coal Co., 365 U. S. 320, 333-335 (1961).

The hospital-Roux arrangement could conceivably have an adverse effect on horizontal competition among anesthesiologists, or among hospitals. Dr. Hyde, who competes with the Roux anesthesiologists, and other hospitals in the area, who compete with East Jefferson, may have grounds to complain that the exclusive contract stifles horizontal competition and therefore has an adverse, albeit indirect, impact on consumer welfare even if it is not a tie.

Exclusive-dealing arrangements may, in some circumstances, create or extend market power of a supplier or the purchaser party to the exclusive-dealing arrangement, and may thus restrain horizontal competition. Exclusive dealing can have adverse economic consequences by allowing one supplier of goods or services unreasonably to deprive other suppliers of a market for their goods, or by allowing one buyer of goods unreasonably to deprive other buyers of a needed source of supply. In determining whether an exclusive-dealing contract is unreasonable, the proper focus is on the structure of the market for the products or services in question — the number of sellers and buyers in the market, the volume of their business, and the ease with which buyers and sellers can redirect their purchases or sales to others. Exclusive dealing is an unreasonable restraint on trade only when a significant fraction of buyers or sellers are frozen out of a market by the exclusive deal. Standard Oil Co. of California v. United States, 337 U. S. 293 (1949). When the sellers of services are numerous and mobile, and the number of buyers is large, exclusive-dealing arrangements of narrow scope pose no threat of adverse economic consequences. To the contrary, they may be substantially procompetitive by ensuring stable markets and encouraging long-term, mutually advantageous business relationships.

At issue here is an exclusive-dealing arrangement between a firm of four anesthesiologists and one relatively small hos*46pital. There is no suggestion that East Jefferson Hospital is likely to create a “bottleneck” in the availability of anesthesiologists that might deprive other hospitals of access to needed anesthesiological services, or that the Roux associates have unreasonably narrowed the range of choices available to other anesthesiologists in search of a hospital or patients that will buy their services. Cf. Associated Press v. United States, 326 U. S. 1 (1945). A firm of four anesthesiologists represents only a very small fraction of the total number of anesthesiologists whose services are available for hire by other hospitals, and East Jefferson is one among numerous hospitals buying such services. Even without engaging in a detailed analysis of the size of the relevant markets we may readily conclude that there is no likelihood that the exclusive-dealing arrangement challenged here will either unreasonably enhance the hospital’s market position relative to other hospitals, or unreasonably permit Roux to acquire power relative to other anesthesiologists. Accordingly, this exclusive-dealing arrangement must be sustained under the rule of reason.

V

For these reasons I conclude that the hospital-Roux contract does not violate § 1 of the Sherman Act. Since anesthesia is a service useful to consumers only when purchased in conjunction with hospital services, the arrangement is not properly characterized as a tie between distinct products. It threatens no additional economic harm to consumers beyond that already made possible by any market power that the hospital may possess. The fact that anesthesia is used only together with other hospital services is sufficient, standing alone, to insulate from attack the hospital’s decision to tie the two types of service.

Whether or not this case involves tying of distinct products, the hospital-Roux contract is subject to scrutiny under the rule of reason as an exclusive-dealing arrangement. Plainly, however, the arrangement forecloses only a small *47fraction of the markets in which anesthesiologists may sell their services, and a still smaller fraction of the market in which hospitals may secure anesthesiological services. The contract therefore survives scrutiny under the rule of reason.

The judgment of the Court of Appeals for the Fifth Circuit should be reversed, and the case should be remanded for any further proceedings on respondent’s remaining claims. See ante, at 5, n. 2.