delivered the opinion of the Court.

This case, like National Bellas Hess, Inc. v. Department of Revenue of Ill., 386 U. S. 753 (1967), involves a State’s attempt to require an out-of-state mail-order house that has neither outlets nor sales representatives in the State to collect and pay a use tax on goods purchased for use within the State. In Bellas Hess we held that a similar Illinois statute violated the Due Process Clause of the Fourteenth Amendment and created an unconstitutional burden on interstate commerce. In particular, we ruled that a “seller whose only connection with customers in the State is by common carrier or the United States mail” lacked the requisite minimum contacts with the State. Id., at 758.

In this case, the Supreme Court of North Dakota declined to follow Bellas Hess because “the tremendous social, economic, commercial, and legal innovations” of the past quarter-century have rendered its holding “obsolete].” 470 N. W. 2d 203, 208 (1991). Having granted certiorari, 502 U. S. 808, we must either reverse the State Supreme Court *302or overrule Bellas Hess. While we agree with much of the state court’s reasoning, we take the former course.

I

Quill is a Delaware corporation with offices and warehouses in Illinois, California, and Georgia. None of its employees work or reside in North Dakota, and its ownership of tangible property in that State is either insignificant or nonexistent.1 Quill sells office equipment and supplies; it solicits business through catalogs and flyers, advertisements in national periodicals, and telephone calls. Its annual national sales exceed $200 million, of which almost $1 million are made to about 3,000 customers in North Dakota. It is the sixth largest vendor of office supplies in the State. It delivers all of its merchandise to its North Dakota customers by mail or common carrier from out-of-state locations.

As a corollary to its sales tax, North Dakota imposes a use tax upon property purchased for storage, use, or consumption within the State. North Dakota requires every “retailer maintaining a place of business in” the State to collect the tax from the consumer and remit it to the State. N. D. Cent. Code §57-40.2-07 (Supp. 1991). In 1987, North Dakota amended the statutory definition of the term “retailer” to include “every person who engages in regular or system*303atic solicitation of a consumer market in th[e] state.” §57-40.2-01(6). State regulations in turn define “regular or systematic solicitation” to mean three or more advertisements within a 12-month period. N. D. Admin. Code § 81-04.1-01-03.1 (1988). Thus, since 1987, mail-order companies that engage in such solicitation have been subject to the tax even if they maintain no property or personnel in North Dakota.

Quill has taken the position that North Dakota does not have the power to compel it to collect a use tax from its North Dakota customers. Consequently, the State, through its Tax Commissioner, filed this action to require Quill to pay taxes (as well as interest and penalties) on all such sales made after July 1, 1987. The trial court ruled in Quill’s favor, finding the case indistinguishable from Bellas Hess; specifically, it found that because the State had not shown that it had spent tax revenues for the benefit of the mail-order business, there was no “nexus to allow the state to define retailer in the manner it chose.” App. to Pet. for Cert. A41.

The North Dakota Supreme Court reversed, concluding that “wholesale changes” in both the economy and the law made it inappropriate to follow Bellas Hess today. 470 N. W. 2d, at 213. The principal economic change noted by the court was the remarkable growth of the mail-order business “from a relatively inconsequential market niche” in 1967 to a “goliath” with annual sales that reached “the staggering figure of $183.3 billion in 1989.” Id., at 208,209. Moreover, the court observed, advances in computer technology greatly eased the burden of compliance with a “‘welter of complicated obligations’ ” imposed by state and local taxing authorities. Id., at 215 (quoting Bellas Hess, 386 U. S., at 759-760).

Equally important, in the court’s view, were the changes in the “legal landscape.” With respect to the Commerce Clause, the court emphasized that Complete Auto Transit, Inc. v. Brady, 430 U. S. 274 (1977), rejected the line of cases holding that the direct taxation of interstate commerce was *304impermissible and adopted instead a “consistent and rational method of inquiry [that focused on] the practical effect of [the] challenged tax.” Mobil Oil Corp. v. Commissioner of Taxes of Vt., 445 U. S. 425, 443 (1980). This and subsequent rulings, the court maintained, indicated that the Commerce Clause no longer mandated the sort of physical-presence nexus suggested in Bellas Hess.

Similarly, with respect to the Due Process Clause, the North Dakota court observed that cases following Bellas Hess had not construed “minimum contacts” to require physical presence within a State as a prerequisite to the legitimate exercise of state power. The state court then concluded that “the Due Process requirement of a ‘minimal connection’ to establish nexus is encompassed within the Complete Auto test” and that the relevant inquiry under the latter test was whether “the state has provided some protection, opportunities, or benefit for which it can expect a return.” 470 N. W. 2d, at 216.

Turning to the case at hand, the state court emphasized that North Dakota had created “an economic climate that fosters demand for” Quill’s products, maintained a legal infrastructure that protected that market, and disposed of 24 tons of catalogs and flyers mailed by Quill into the State every year. Id., at 218-219. Based on these facts, the court concluded that Quill’s “economic presence” in North Dakota depended on services and benefits provided by the State and therefore generated “a constitutionally sufficient nexus to justify imposition of the purely administrative duty of collecting and remitting the use tax.” Id., at 219.2

*305II

As in a number of other cases involving the application of state taxing statutes to out-of-state sellers, our holding in Bellas Hess relied on both the Due Process Clause and the Commerce Clause. Although the “two claims are closely related,” Bellas Hess, 386 U. S., at 756, the Clauses pose distinct limits on the taxing powers of the States. Accordingly, while a State may, consistent with the Due Process Clause, have the authority to tax a particular taxpayer, imposition of the tax may nonetheless violate the Commerce Clause. See, e. g., Tyler Pipe Industries, Inc. v. Washington State Dept. of Revenue, 483 U. S. 232 (1987).

The two constitutional requirements differ fundamentally, in several ways. As discussed at greater length below, see Part IV, infra, the Due Process Clause and the Commerce Clause reflect different constitutional concerns. Moreover, while Congress has plenary power to regulate commerce among the States and thus may authorize state actions that burden interstate commerce, see International Shoe Co. v. Washington, 326 U. S. 310, 315 (1945), it does not similarly have the power to authorize violations of the Due Process Clause.

Thus, although we have not always been precise in distinguishing between the two, the Due Process Clause and the Commerce Clause are analytically distinct.

“ ‘Due process’ and ‘commerce clause’ conceptions are not always sharply separable in dealing with these problems. ... To some extent they overlap. If there is a want of due process to sustain the tax, by that fact alone any burden the tax imposes on the commerce among the states becomes ‘undue.’ But, though overlapping, the two conceptions are not identical. There may be more than sufficient factual connections, with economic and legal effects, between the transaction and the taxing state to sustain the tax as against due process *306objections. Yet it may fall because of its burdening effect upon the commerce. And, although the two notions cannot always be separated, clarity of consideration and of decision would be promoted if the two issues are approached, where they are presented, at least tentatively as if they were separate and distinct, not intermingled ones.” International Harvester Co. v. Department of Treasury, 322 U. S. 340, 353 (1944) (Rutledge, J., concurring in part and dissenting in part).

Heeding Justice Rutledge’s counsel, we consider each constitutional limit in turn.

Ill

The Due Process Clause “requires some definite link, some minimum connection, between a state and the person, property or transaction it seeks to tax,” Miller Brothers Co. v. Maryland, 347 U. S. 340, 344-345 (1954), and that the “income attributed to the State for tax purposes must be rationally related to ‘values connected with the taxing State,’” Moorman Mfg. Co. v. Bair, 437 U. S. 267, 273 (1978) (citation omitted). Here, we are concerned primarily with the first of these requirements. Prior to Bellas Hess, we had held that that requirement was satisfied in a variety of circumstances involving use taxes. For example, the presence of sales personnel in the State3 or the maintenance of local retail stores in the State4 justified the exercise of that power because the seller’s local activities were “plainly accorded the protection and services of the taxing State.” Bellas Hess, 386 U. S., at 757. The furthest extension of that power was recognized in Scripto, Inc. v. Carson, 362 U. S. 207 (1960), in which the Court upheld a use tax despite the fact that all of the seller’s in-state solicitation was performed by independent contractors. These cases all involved some sort of physical presence within the State, and in Bellas Hess *307the Court suggested that such presence was not only sufficient for jurisdiction under the Due Process Clause, but also necessary. We expressly declined to obliterate the “sharp distinction ... between mail-order sellers with retail outlets, solicitors, or property within a State, and those who do no more than communicate with customers in the State by mail or common carrier as a part of a general interstate business.” 386 U. S., at 758.

Our due process jurisprudence has evolved substantially in the 25 years since Bellas Hess, particularly in the area of judicial jurisdiction. Building on the seminal case of International Shoe Co. v. Washington, 326 U. S. 310 (1945), we have framed the relevant inquiry as whether a defendant had minimum contacts with the jurisdiction “such that the maintenance of the suit does not offend ‘traditional notions of fair play and substantial justice.’” Id., at 316 (quoting Milliken v. Meyer, 311 U. S. 457, 463 (1940)). In that spirit, we have abandoned more formalistic tests that focused on a defendant’s “presence” within a State in favor of a more flexible inquiry into whether a defendant’s contacts with the forum made it reasonable, in the context of our federal system of Government, to require it to defend the suit in that State. In Shaffer v. Heitner, 433 U. S. 186, 212 (1977), the Court extended the flexible approach that International Shoe had prescribed for purposes of in personam jurisdiction to in rem jurisdiction, concluding that “all assertions of state-court jurisdiction must be evaluated according to the standards set forth in International Shoe and its progeny.”

Applying these principles, we have held that if a foreign corporation purposefully avails itself of the benefits of an economic market in the forum State, it may subject itself to the State’s in personam jurisdiction even if it has no physical presence in the State. As we explained in Burger King Corp. v. Rudzewicz, 471 U. S. 462 (1985):

“Jurisdiction in these circumstances may not be avoided merely because the defendant did not physi*308cally enter the forum State. Although territorial presence frequently will enhance a potential defendant’s affiliation with a State and reinforce the reasonable foreseeability of suit there, it is an inescapable fact of modern commercial life that a substantial amount of business is transacted solely by mail and wire communications across state lines, thus obviating the need for physical presence within a State in which business is conducted. So long as a commercial actor’s efforts are ‘purposefully directed’ toward residents of another State, we have consistently rejected the notion that an absence of physical contacts can defeat personal jurisdiction there.” Id., at 476 (emphasis in original).

Comparable reasoning justifies the imposition of the collection duty on a mail-order house that is engaged in continuous and widespread solicitation of business within a State. Such a corporation clearly has “fair warning that [its] activity may subject [it] to the jurisdiction of a foreign sovereign.” Shaffer v. Heitner, 433 U. S., at 218 (Stevens, J., concurring in judgment). In “modern commercial life” it matters little that such solicitation is accomplished by a deluge of catalogs rather than a phalanx of drummers: The requirements of due process are met irrespective of a corporation’s lack of physical presence in the taxing State. Thus, to the extent that our decisions have indicated that the Due Process Clause requires physical presence in a State for the imposition of duty to collect a use tax, we overrule those holdings as superseded by developments in the law of due process.

In this case, there is no question that Quill has purposefully directed its activities at North Dakota residents, that the magnitude of those contacts is more than sufficient for due process purposes, and that the use tax is related to the benefits Quill receives from access to the State. We therefore agree with the North Dakota Supreme Court’s conclusion that the Due Process Clause does not bar enforcement of that State’s use tax against Quill.

*309HH <J

Article I, § 8, cl. 3, of the Constitution expressly authorizes Congress to “regulate Commerce with foreign Nations, and among the several States.” It says nothing about the protection of interstate commerce in the absence of any action by Congress. Nevertheless, as Justice Johnson suggested in his concurring opinion in Gibbons v. Ogden, 9 Wheat. 1, 231-232, 239 (1824), the Commerce Clause is more than an affirmative grant of power; it has a negative sweep as well. The Clause, in Justice Stone’s phrasing, “by its own force” prohibits certain state actions that interfere with interstate commerce. South Carolina State Highway Dept. v. Barnwell Brothers, Inc., 303 U. S. 177, 185 (1938).

Our interpretation of the “negative” or “dormant” Commerce Clause has evolved substantially over the years, particularly as that Clause concerns limitations on state taxation powers. See generally P. Hartman, Federal Limitations on State and Local Taxation §§2:9-2:17 (1981). Our early cases, beginning with Brown v. Maryland, 12 Wheat. 419 (1827), swept broadly, and in Leloup v. Port of Mobile, 127 U. S. 640, 648 (1888), we declared that “no State has the right to lay a tax on interstate commerce in any form.” We later narrowed that rule and distinguished between direct burdens on interstate commerce, which were prohibited, and indirect burdens, which generally were not. See, e. g., Sanford v. Poe, 69 F. 546 (CA6 1895), aff’d sub nom. Adams Express Co. v. Ohio State Auditor, 165 U. S. 194, 220 (1897). Western Live Stock v. Bureau of Revenue, 303 U. S. 250, 256-258 (1938), and subsequent decisions rejected this formal, categorical analysis and adopted a “multiple-taxation doctrine” that focused not on whether a tax was “direct” or “indirect” but rather on whether a tax subjected interstate commerce to a risk of multiple taxation. However, in Freeman v. Hewit, 329 U. S. 249, 256 (1946), we embraced again the formal distinction between direct and indirect taxation, invalidating Indiana’s imposition of a gross receipts tax on a *310particular transaction because that application would “im-poste] a direct tax on interstate sales.” Most recently, in Complete Auto Transit, Inc. v. Brady, 430 U. S., at 285, we renounced the Freeman approach as “attaching constitutional significance to a semantic difference.” We expressly overruled one of Freeman’s progeny, Spector Motor Service, Inc. v. O’Connor, 340 U. S. 602 (1951), which held that a tax on “the privilege of doing interstate business” was unconstitutional, while recognizing that a differently denominated tax with the same economic effect would not be unconstitutional. Spector, as we observed in Railway Express Agency, Inc. v. Virginia, 358 U. S. 434, 441 (1959), created a situation in which “magic words or labels” could “disable an otherwise constitutional levy.” Complete Auto emphasized the importance of looking past “the formal language of the tax statute [to] its practical effect,” 430 U. S., at 279, and set forth a four-part test that continues to govern the validity of state taxes under the Commerce Clause.5

Bellas Hess was decided in 1967, in the middle of this latest rally between formalism and pragmatism. Contrary to the suggestion of the North Dakota Supreme Court, this timing does not mean that Complete Auto rendered Bellas Hess “obsolete.” Complete Auto rejected Freeman and Spector’s formal distinction between “direct” and “indirect” taxes on interstate commerce because that formalism allowed the validity of statutes to hinge on “legal terminology,” “draftsmanship and phraseology.” 430 U. S., at 281. Bellas Hess *311did not rely on any such labeling of taxes and therefore did not automatically fall with Freeman and its progeny.

While contemporary Commerce Clause jurisprudence might not dictate the same result were the issue to arise for the first time today, Bellas Hess is not inconsistent with Complete Auto and our recent cases. Under Complete Auto’s four-part test, we will sustain a tax against a Commerce Clause challenge so long as the “tax [1] is applied to an activity with a substantial nexus with the taxing State, [2] is fairly apportioned, [3] does not discriminate against interstate commerce, and [4] is fairly related to the services provided by the State.” 430 U. S., at 279. Bellas Hess concerns the first of these tests and stands for the proposition that a vendor whose only contacts with the taxing State are by mail or common carrier lacks the “substantial nexus” required by the Commerce Clause.

Thus, three weeks after Complete Auto was handed down, we cited Bellas Hess for this proposition and discussed the case at some length. In National Geographic Society v. California Bd. of Equalization, 430 U. S. 551, 559 (1977), we affirmed the continuing vitality of Bellas Hess’ “sharp distinction . . . between mail-order sellers with [a physical presence in the taxing] State and those ... who do no more than communicate with customers in the State by mail or common carrier as part of a general interstate business.” We have continued to cite Bellas Hess with approval ever since. For example, in Goldberg v. Sweet, 488 U. S. 252, 263 (1989), we expressed “doubt that termination of an interstate telephone call, by itself, provides a substantial enough nexus for a State to tax a call. See National Bellas Hess . . . (receipt of mail provides insufficient nexus).” See also D. H. Holmes Co. v. McNamara, 486 U. S. 24, 33 (1988); Commonwealth Edison Co. v. Montana, 453 U. S. 609, 626 (1981); Mobil Oil Corp. v. Commissioner of Taxes, 445 U. S., at 437; National Geographic Society, 430 U. S., at 559. For these reasons, we disagree with the State Supreme Court’s conclu*312sion that our decision in Complete Auto undercut the Bellas Hess rule.

The State of North Dakota relies less on Complete Auto and more on the evolution of our due process jurisprudence. The State contends that the nexus requirements imposed by the Due Process and Commerce Clauses are equivalent and that if, as we concluded above, a mail-order house that lacks a physical presence in the taxing State nonetheless satisfies the due process “minimum contacts” test, then that corporation also meets the Commerce Clause “substantial nexus” test. We disagree. Despite the similarity in phrasing, the nexus requirements of the Due Process and Commerce Clauses are not identical. The two standards are animated by different constitutional concerns and policies.

Due process centrally concerns the fundamental fairness of governmental activity. Thus, at the most general level, the due process nexus analysis requires that we ask whether an individual’s connections with a State are substantial enough to legitimate the State’s exercise of power over him. We have, therefore, often identified “notice” or “fair warning” as the analytic touchstone of due process nexus analysis. In contrast, the Commerce Clause and its nexus requirement are informed not so much by concerns about fairness for the individual defendant as by structural concerns about the effects of state regulation on the national economy. Under the Articles of Confederation, state taxes and duties hindered and suppressed interstate commerce; the Framers intended the Commerce Clause as a cure for these structural ills. See generally The Federalist Nos. 7, 11 (A. Hamilton). It is in this light that we have interpreted the negative implication of the Commerce Clause. Accordingly, we have ruled that that Clause prohibits discrimination against interstate commerce, see, e. g., Philadelphia v. New Jersey, 437 U. S. 617 (1978), and bars state regulations that unduly burden interstate commerce, see, e. g., Kassel v. Consolidated Freightways Corp. of Del., 450 U. S. 662 (1981).

*313The Complete Auto analysis reflects these concerns about the national economy. The second and third parts of that analysis, which require fair apportionment and nondiscrimination, prohibit taxes that pass an unfair share of the tax burden onto interstate commerce. The first and fourth prongs, which require a substantial nexus and a relationship between the tax and state-provided services, limit the reach of state taxing authority so as to ensure that state taxation does not unduly burden interstate commerce.6 Thus, the “substantial nexus” requirement is not, like due process’ “minimum contacts” requirement, a proxy for notice, but rather' a means for limiting state burdens on interstate commerce. Accordingly, contrary to the State’s suggestion, a corporation may have the “minimum contacts” with a taxing State as required by the Due Process Clause, and yet lack the “substantial nexus” with that State as required by the Commerce Clause.7

*314The State Supreme Court reviewed our recent Commerce Clause decisions and concluded that those rulings signaled a “retreat from the formalistic constrictions of a stringent physical presence test in favor of a more flexible substantive approach” and thus supported its decision not to apply Bellas Hess. 470 N. W. 2d, at 214 (citing Standard Pressed Steel Co. v. Department of Revenue of Wash., 419 U. S. 560 (1975), and Tyler Pipe Industries, Inc. v. Washington State Dept. of Revenue, 483 U. S. 232 (1987)). Although we agree with the state court’s assessment of the evolution of our cases, we do not share its conclusion that this evolution indicates that the Commerce Clause ruling of Bellas Hess is no longer good law.

First, as the state court itself noted, 470 N. W. 2d, at 214, all of these cases involved taxpayers who had a physical presence in the taxing State and therefore do not directly conflict with the rule of Bellas Hess or compel that it be overruled. Second, and more importantly, although our Commerce Clause jurisprudence now favors more flexible balancing analyses, we have never intimated a desire to reject all established “bright-line” tests. Although we have not, in our review of other types of taxes, articulated the same physical-presence requirement that Bellas Hess established for sales and use taxes, that silence does not imply repudiation of the Bellas Hess rule.

Complete Auto, it is true, renounced Freeman and its progeny as “formalistic.” But not all formalism is alike. Spector’s formal distinction between taxes on the “privilege of doing business” and all other taxes served no purpose within our Commerce Clause jurisprudence, but stood “only as a trap for the unwary draftsman.” Complete Auto, 430 U. S., at 279. In contrast, the bright-line rule of Bellas Hess furthers the ends of the dormant Commerce Clause. Undue *315burdens on interstate commerce may be avoided not only by a case-by-case evaluation of the actual burdens imposed by particular regulations or taxes, but also, in some situations, by the demarcation of a discrete realm of commercial activity that is free from interstate taxation. Bellas Hess followed the latter approach and created a safe harbor for vendors “whose only connection with customers in the [taxing] State is by common carrier or the United States mail.” Under Bellas Hess, such vendors are free from state-imposed duties to collect sales and use taxes.8

Like other bright-line tests, the Bellas Hess rule appears artificial at its edges: Whether or not a State may compel a vendor to collect a sales or use tax may turn on the presence in the taxing State of a small sales force, plant, or office. Cf. National Geographic Society v. California Bd. of Equalization, 430 U. S. 551 (1977); Scripto, Inc. v. Carson, 362 U. S. 207 (1960). This artificiality, however, is more than offset by the benefits of a clear rule. Such a rule firmly establishes the boundaries of legitimate state authority to impose a duty to collect sales and use taxes and reduces litigation concerning those taxes. This benefit is important, for as we have so frequently noted, our law in this area is something of a “quagmire” and the “application of constitutional principles to specific state statutes leaves much room for controversy and confusion and little in the way of precise guides to the States in the exercise of their indispensable power of *316taxation.” Northwestern States Portland Cement Co. v. Minnesota, 358 U. S. 450, 457-458 (1959).

Moreover, a bright-line rule in the area of sales and use taxes also encourages settled expectations and, in doing so, fosters investment by businesses and individuals.9 Indeed, it is not unlikely that the mail-order industry’s dramatic growth over the last quarter century is due in part to the bright-line exemption from state taxation created in Bellas Hess.

Notwithstanding the benefits of bright-line tests, we have, in some situations, decided to replace such tests with more contextual balancing inquiries. For example, in Arkansas Electric Cooperative Corp. v. Arkansas Pub. Serv. Comm’n, 461 U. S. 375 (1983), we reconsidered a bright-line test set forth in Public Util. Comm’n of R. I. v. Attleboro Steam & Electric Co., 273 U. S. 83 (1927). Attleboro distinguished between state regulation of wholesale sales of electricity, which was constitutional as an “indirect” regulation of interstate commerce, and state regulation of retail sales of electricity, which was unconstitutional as a “direct regulation” of commerce. In Arkansas Electric, we considered whether to *317“follow the mechanical test set out in Attleboro, or the balance-of-interests test applied in our Commerce Clause cases.” 461 U. S., at 390-391. We first observed that “the principle of stare decisis counsels us, here as elsewhere, not lightly to set aside specific guidance of the sort we find in Attleboro.” Id., at 391. In deciding to reject the Attleboro analysis, we were influenced by the fact that the “mechanical test” was “anachronistic,” that the Court had rarely relied on the test, and that we could “see no strong reliance interests” that would be upset by the rejection of that test. 461 U. S., at 391-392. None of those factors obtains in this case. First, the Attleboro rule was “anachronistic” because it relied on formal distinctions between “direct” and “indirect” regulation (and on the regulatory counterparts of our Freeman line of cases); as discussed above, Bellas Hess turned on a different logic and thus remained sound after the Court repudiated an analogous distinction in Complete Auto. Second, unlike the Attleboro rule, we have, in our decisions, frequently relied on the Bellas Hess rule in the last 26 years, see supra, at 311, and we have never intimated in our review of sales or use taxes that Bellas Hess was unsound. Finally, again unlike the Attleboro rule, the Bellas Hess rule has engendered substantial reliance and has become part of the basic framework of a sizable industry. The “interest in stability and orderly development of the law” that undergirds the doctrine of stare decisis, see Runyon v. McCrary, 427 U. S. 160, 190-191 (1976) (Stevens, J., concurring), therefore counsels adherence to settled precedent.

In sum, although in our cases subsequent to Bellas Hess and concerning other types of taxes we have not adopted a similar bright-line, physical-presence requirement, our reasoning in those cases does not compel that we now reject the rule that Bellas Hess established in the area of sales and use taxes. To the contrary, the continuing value of a bright-line rule in this area and the doctrine and principles of stare deci-sis indicate that the Bellas Hess rule remains good law. For *318these reasons, we disagree with the North Dakota Supreme Court’s conclusion that the time has come to renounce the bright-line test of Bellas Hess.

This aspect of our decision is made easier by the fact that the underlying issue is not only one that Congress may be better qualified to resolve,10 but also one that Congress has the ultimate power to resolve. No matter how we evaluate the burdens that use taxes impose on interstate commerce, Congress remains free to disagree with our conclusions. See Prudential Insurance Co. v. Benjamin, 328 U. S. 408 (1946). Indeed, in recent years Congress has considered legislation that would “overrule” the Bellas Hess rule.11 Its decision not to take action in this direction may, of course, have been dictated by respect for our holding in Bellas Hess that the Due Process Clause prohibits States from imposing such taxes, but today we have put that problem to rest. Accordingly, Congress is now free to decide whether, when, and to what extent the States may burden interstate mail-order concerns with a duty to collect use taxes.

Indeed, even if we were convinced that Bellas Hess was inconsistent with our Commerce Clause jurisprudence, “this very fact [might] giv[e us] pause and counse[l] withholding our hand, at least for now. Congress has the power to protect interstate commerce from intolerable or even undesirable burdens.” Commonwealth Edison Co. v. Montana, 463 U. S., at 637 (White, J., concurring). In this situation, it *319may be that “the better part of both wisdom and valor is to respect the judgment of the other branches of the Government.” Id., at 638.

The judgment of the Supreme Court of North Dakota is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

with whom Justice Kennedy and Justice Thomas join, concurring in part and concurring in the judgment.

National Bellas Hess, Inc. v. Department of Revenue of Ill., 386 U. S. 763 (1967), held that the Due Process and Commerce Clauses of the Constitution prohibit a State from imposing the duty of use-tax collection and payment upon a seller whose only connection with the State is through common carrier or the United States mail. I agree with the Court that the Due Process Clause holding of Bellas Hess should be overruled. Even before Bellas Hess, we had held, correctly I think, that state regulatory jurisdiction could be asserted on the basis of contacts with the State through the United States mail. See Travelers Health Assn. v. Virginia ex rel. State Corp. Comm’n, 339 U. S. 643, 646-650 (1950) (blue sky laws). It is difficult to discern any principled basis for distinguishing between jurisdiction to regulate and jurisdiction to tax. As an original matter, it might have been possible to distinguish between jurisdiction to tax and jurisdiction to compel collection of taxes as agent for the State, but we have rejected that. National Geographic Society v. California Bd. of Equalization, 430 U. S. 551, 558 (1977); Scripto, Inc. v. Carson, 362 U. S. 207, 211 (1960). I agree with the Court, moreover, that abandonment of Bellas Hess’ due process holding is compelled by reasoning “ [comparable” to that contained in our post-1967 cases dealing with state jurisdiction to adjudicate. Ante; at 308. I do not understand this to mean that the due process standards for *320adjudicative jurisdiction and those for legislative (or prescriptive) jurisdiction are necessarily identical; and on that basis I join Parts I, II, and III of the Court’s opinion. Compare Asahi Metal Industry Co. v. Superior Court of Cal., Solano Cty., 480 U. S. 102 (1987), with American Oil Co. v. Neill, 380 U. S. 451 (1965).

I also agree that the Commerce Clause holding of Bellas Hess should not be overruled. Unlike the Court, however, I would not revisit the merits of that holding, but would adhere to it on the basis of stare decisis. American Trucking Assns., Inc. v. Smith, 496 U. S. 167, 204 (1990) (Scalia, J., concurring in judgment). Congress has the final say over regulation of interstate commerce, and it can change the rule of Bellas Hess by simply saying so. We have long recognized that the doctrine of stare decisis has “special force” where “Congress remains free to alter what we have done.” Patterson v. McLean Credit Union, 491 U. S. 164, 172-173 (1989). See also Hilton v. South Carolina Public Railways Comm’n, 502 U. S. 197, 202 (1991); Illinois Brick Co. v. Illinois, 431 U. S. 720, 736 (1977). Moreover, the demands of the doctrine are “at their acme . . . where reliance interests are involved.” Payne v. Tennessee, 501 U. S. 808, 828 (1991). As the Court notes, “the Bellas Hess rule has engendered substantial reliance and has become part of the basic framework of a sizable industry.” Ante, at 317.

I do not share Justice White’s view that we may disregard these reliance interests because it has become unreasonable to rely upon Bellas Hess. Post, at 331-332. Even assuming for the sake of argument (I do not consider the point) that later decisions in related areas are inconsistent with the principles upon which Bellas Hess rested, we have never acknowledged that, but have instead carefully distinguished the case on its facts. See, e. g., D. H. Holmes Co. v. McNamara, 486 U. S. 24, 33 (1988); National Geographic Society, supra, at 559. It seems to me important that we retain our ability — and, what comes to the same thing, that *321we maintain public confidence in our ability — sometimes to adopt new principles for the resolution of new issues without abandoning clear holdings of the past that those principles contradict. We seemed to be doing that in this area. Having affirmatively suggested that the “physical presence” rule could be reconciled with our new jurisprudence, we ought not visit economic hardship upon those who took us at our word. We have recently told lower courts that “[i]f a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, [they] should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions.” Rodriguez de Quijas v. Shearson/American Express, Inc., 490 U. S. 477, 484 (1989). It is strangely incompatible with this to demand that private parties anticipate our overrulings. It is my view, in short, that reliance upon a square, unabandoned holding of the Supreme Court is always justifiable reliance (though reliance alone may not always carry the day). Finally, the “physical presence” rule established in Bellas Hess is not “unworkable,” Patterson, supra, at 173; to the contrary, whatever else may be the substantive pros and cons of the rule, the “bright-line” regime that it establishes, see ante, at 314, is unqualifiedly in its favor. Justice White’s concern that reaffirmance of Bellas Hess will lead to a flurry of litigation over the meaning of “physical presence,” see post, at 331, seems to me contradicted by 25 years of experience under the decision.

For these reasons, I concur in the judgment of the Court and join Parts I, II, and III of its opinion.

concurring in part and dissenting in part.

Today the Court repudiates that aspect of our decision in National Bellas Hess, Inc. v. Department of Revenue of Ill., 386 U. S. 753 (1967), which restricts, under the Due Process Clause of the Fourteenth Amendment, the power of the States to impose use tax collection responsibilities on out-*322of-state mail-order businesses that do not have a “physical presence” in the State. The Court stops short, however, of giving Bellas Hess the complete burial it justly deserves. In my view, the Court should also overrule that part of Bellas Hess which justifies its holding under the Commerce Clause. I, therefore, respectfully dissent from Part IV.

M

In Part IV of its opinion, the majority goes to some lengths to justify the Bellas Hess physical-presence requirement under our Commerce Clause jurisprudence. I am unpersuaded by its interpretation of our cases. In Bellas Hess, the majority placed great weight on the interstate quality of the mail-order sales, stating that “it is difficult to conceive of commercial transactions more exclusively interstate in character than the mail-order transactions here involved.” Id., at 759. As the majority correctly observes, the idea of prohibiting States from taxing “exclusively interstate” transactions had been an important part of our jurisprudence for many decades, ranging intermittently from such cases as Case of State Freight Tax, 15 Wall. 232, 279 (1873), through Freeman v. Hewit, 329 U. S. 249, 256 (1946), and Spector Motor Service, Inc. v. O’Connor, 340 U. S. 602 (1951). But though it recognizes that Bellas Hess was decided amidst an upheaval in our Commerce Clause jurisprudence, in which we began to hold that “a State, with proper drafting, may tax exclusively interstate commerce so long as the tax does not create any effect forbidden by the Commerce Clause,” Complete Auto Transit, Inc. v. Brady, 430 U. S. 274, 285 (1977), the majority draws entirely the wrong conclusion from this period of ferment.

The Court attempts to paint Bellas Hess in a different hue from Freeman and Spector because the former “did not rely” on labeling taxes that had “direct” and “indirect” effects on interstate commerce. See ante, at 310. Thus, the Court concludes, Bellas Hess “did not automatically fall with Free*323man and its progeny” in our decision in Complete Auto. See ante, at 311. I am unpersuaded by this attempt to distinguish Bellas Hess from Freeman and Spector, both of which were repudiated by this Court. See Complete Auto, supra, at 288-289, and n. 15. What we disavowed in Complete Auto was not just the “formal distinction between ‘direct’ and ‘indirect’ taxes on interstate commerce,” ante, at 310, but also the whole notion underlying the Bellas Hess physical-presence rule — that “interstate commerce is immune from state taxation,” Complete Auto, supra, at 288.

The Court compounds its misreading by attempting to show that Bellas Hess “is not inconsistent with Complete Auto and our recent cases.” Ante, at 311. This will be news to commentators, who have rightly criticized Bellas Hess.1 Indeed, the majority displays no small amount of audacity in claiming that our decision in National Geographic Society v. California Bd. of Equalization, 430 U. S. 551, 559 (1977), which was rendered several weeks after Complete Auto, reaffirmed the continuing vitality of Bellas Hess. See ante, at 311.

Our decision in that case did just the opposite. National Geographic held that the National Geographic Society was liable for use tax collection responsibilities in California. The Society conducted an out-of-state mail-order business similar to the one at issue here and in Bellas Hess, and in addition, maintained two small offices in California that solicited advertisements for National Geographic Magazine. The Society argued that its physical presence in California was unrelated to its mail-order sales, and thus that the Bel-*324las Hess rule compelled us to hold that the tax collection responsibilities could not be imposed. We expressly rejected that view, holding that the “requisite nexus for requiring an out-of-state seller [the Society] to collect and pay the use tax is not whether the duty to collect the use tax relates to the seller’s activities carried on within the State, but simply whether the facts demonstrate ‘some definite link, some minimum connection, between (the State and) the person... it seeks to tax.’” 430 U. S., at 561 (citation omitted).

By decoupling any notion of a transactional nexus from the inquiry, the National Geographic Court in fact repudiated the free trade rationale of the Bellas Hess majority. Instead, the National Geographic Court relied on a due process-type minimum contacts analysis that examined whether a link existed between the seller and the State wholly apart from the seller’s in-state transaction that was being taxed. Citations to Bellas Hess notwithstanding, see 430 U. S., at 559, it is clear that rather than adopting the rationale of Bellas Hess, the National Geographic Court was instead politely brushing it aside. Even were I to agree that the free trade rationale embodied in Bellas Hess’ rule against taxes of purely interstate sales was required by our cases prior to 1967, therefore, I see no basis in the majority’s opening premise that this substantive underpinning of Bellas Hess has not since been disavowed by our cases.2

*325II

The Court next launches into an uncharted and treacherous foray into differentiating between the “nexus” requirements under the Due Process and Commerce Clauses. As the Court explains: “Despite the similarity in phrasing, the nexus requirements of the Due Process and Commerce Clauses are not identical. The two standards are animated by different constitutional concerns and policies.” Ante, at 312. The due process nexus, which the Court properly holds is met in this case, see ante, at Part III, “concerns the fundamental fairness of governmental activity.” Ante, at 312. The Commerce Clause nexus requirement, on the other hand, is “informed not so much by concerns about fairness for the individual defendant as by structural concerns about the effects of state regulation on the national economy.” Ibid.

Citing Complete Auto, the Court then explains that the Commerce Clause nexus requirement is not “like due process’ ‘minimum contacts’ requirement, a proxy for notice, but rather a means for limiting state burdens on interstate commerce.” Ante, at 313. This is very curious, because parts two and three of the Complete Auto test, which require fair apportionment and nondiscrimination in order that interstate commerce not be unduly burdened, now appear to become the animating features of the nexus requirement, which is the first prong of the Complete Auto inquiry. The Court freely acknowledges that there is no authority for this novel interpretation of our cases and that we have never before found, as we do in this case, sufficient contacts for due process purposes but an insufficient nexus under the Commerce Clause. See ante, at 313-314, and n. 6.

The majority’s attempt to disavow language in our opinions acknowledging the presence of due process require*326ments in the Complete Auto test is also unpersuasive. See ante, at 313-314, n. 7 (citing Trinova Corp. v. Michigan Dept. of Treasury, 498 U. S. 358, 373 (1991)). Instead of explaining the doctrinal origins of the Commerce Clause nexus requirement, the majority breezily announces the rule and moves on to other matters. See ante, at 313-314. In my view, before resting on the assertion that the Constitution mandates inquiry into two readily distinct “nexus” requirements, it would seem prudent to discern the origins of the “nexus” requirement in order better to understand whether the Court’s concern traditionally has been with the fairness of a State’s tax or some other value.

The cases from which the Complete Auto Court derived the nexus requirement in its four-part test convince me that the issue of “nexus” is really a due process fairness inquiry. In explaining the sources of the four-part inquiry in Complete Auto, the Court relied heavily on Justice Rutledge’s separate concurring opinion in Freeman v. Hewit, 329 U. S. 249 (1946), the case whose majority opinion the Complete Auto Court was in the process of comprehensively disavowing. Instead of the formalistic inquiry into whether the State was taxing interstate commerce, the Complete Auto Court adopted the more functionalist approach of Justice Rutledge in Freeman. See Complete Auto, 430 U. S., at 280-281. In conducting his inquiry, Justice Rutledge used language that by now should be familiar, arguing that a tax was unconstitutional if the activity lacked a sufficient connection to the State to give “jurisdiction to tax,” Freeman, supra, at 271; or if the tax discriminated against interstate commerce; or if the activity was subjected to multiple tax burdens. 329 U. S., at 276-277. Justice Rutledge later refined these principles in Memphis Natural Gas Co. v. Stone, 335 U. S. 80 (1948), in which he described the principles that the Complete Auto Court would later substantially adopt: “[I]t is enough for me to sustain the tax imposed in this case that it is one clearly within the state’s power to lay insofar *327as any limitation of due process or ‘jurisdiction to tax’ in that sense is concerned; it is nondiscriminatory . . . ; [it] is duly apportioned ...; and cannot be repeated by any other state.” 335 U. S., at 96-97 (concurring opinion) (footnotes omitted).

By the time the Court decided Northwestern States Portland Cement Co. v. Minnesota, 358 U. S. 450 (1959), Justice Rutledge was no longer on the Court, but his view of the nexus requirement as grounded in the Due Process Clause was decisively adopted. In rejecting challenges to a state tax based on the Due Process and Commerce Clauses, the Court stated: “The taxes imposed are levied only on that portion of the taxpayer’s net income which arises from its activities within the taxing State. These activities form a sufficient ‘nexus between such a tax and transactions within a state for which the tax is an exaction.’ ” Id., at 464 (citation omitted). The Court went on to observe that “[i]t strains reality to say, in terms of our decisions, that each of the corporations here was not sufficiently involved in local events to forge ‘some definite link, some minimum connection’ sufficient to satisfy due process requirements.” Id., at 464-465 (quoting Miller Brothers Co. v. Maryland, 347 U. S. 340, 344-345 (1954)). When the Court announced its four-part synthesis in Complete Auto, the nexus requirement was definitely traceable to concerns grounded in the Due Process Clause, and not the Commerce Clause, as the Court’s discussion of the doctrinal antecedents for its rule made clear. See Complete Auto, supra, at 281-282, 285. For the Court now to assert that our Commerce Clause jurisprudence supports a separate notion of nexus is without precedent or explanation.

Even were there to be such an independent requirement under the Commerce Clause, there is no relationship between the physical-presence/nexus rule the Court retains and Commerce Clause considerations that allegedly justify it. Perhaps long ago a seller’s “physical presence” was a sufficient part of a trade to condition imposition of a tax on *328such presence. But in today’s economy, physical presence frequently has very little to do with a transaction a State might seek to tax. Wire transfers of money involving billions of dollars occur every day; purchasers place orders with sellers by fax, phone, and computer linkup; sellers ship goods by air, road, and sea through sundry delivery services without leaving their place of business. It is certainly true that the days of the door-to-door salesperson are not gone. Nevertheless, an out-of-state direct marketer derives numerous commercial benefits from the State in which it does business. These advantages include laws establishing sound local banking institutions to support credit transactions; courts to ensure collection of the purchase price from the seller’s customers; means of waste disposal from garbage generated by mail-order solicitations; and creation and enforcement of consumer protection laws, which protect buyers and sellers alike, the former by ensuring that they will have a ready means of protecting against fraud, and the latter by creating a climate of consumer confidence that inures to the benefit of reputable dealers in mail-order transactions. To create, for the first time, a nexus requirement under the Commerce Clause independent of that established for due process purposes is one thing; to attempt to justify an anachronistic notion of physical presence in economic terms is quite another.

Ill

The illogic of retaining the physical-presence requirement in these circumstances is palpable. Under the majority’s analysis, and our decision in National Geographic, an out-of-state seller with one salesperson in a State would be subject to use tax collection burdens on its entire mail-order sales even if those sales were unrelated to the salesperson’s solicitation efforts. By contrast, an out-of-state seller in a neighboring State could be the dominant business in the putative taxing State, creating the greatest infrastructure burdens and undercutting the State’s home companies by its compara*329tive price advantage in selling products free of use taxes, and yet not have to collect such taxes if it lacks a physical presence in the taxing State. The majority clings to the physical-presence rule not because of any logical relation to fairness or any economic rationale related to principles underlying the Commerce Clause, but simply out of the supposed convenience of having a bright-line rule. I am less impressed by the convenience of such adherence than the unfairness it produces. Here, convenience should give way. Cf. Complete Auto, supra, at 289, n. 15 (“We believe, however, that administrative convenience ... is insufficient justification for abandoning the principle that 'interstate commerce may be made to pay its way’ ”).

Also very questionable is the rationality of perpetuating a rule that creates an interstate tax shelter for one form of business — mail-order sellers — but no countervailing advantage for its competitors. If the Commerce Clause was intended to put businesses on an even playing field, the majority’s rule is hardly a way to achieve that goal. Indeed, arguably even under the majority’s explanation for its “Commerce Clause nexus” requirement, the unfairness of its rule on retailers other than direct marketers should be taken into account. See ante, at 312 (stating that the Commerce Clause nexus requirement addresses the “structural concerns about the effects of state regulation on the national economy”). I would think that protectionist rules favoring a $180-billion-a-year industry might come within the scope of such “structural concerns.” See Brief for State of New Jersey as Amicus Curiae 4.

IV

The Court attempts to justify what it rightly acknowledges is an “artificial” rule in several ways. See ante, at 315. First, it asserts that the Bellas Hess principle “firmly establishes the boundaries of legitimate state authority to impose a duty to collect sales and use taxes and reduces litigation concerning those taxes.” Ante, at 315. It is very doubtful, *330however, that the Court’s opinion can achieve its aims. Certainly our cases now demonstrate two “bright-line” rules for mail-order sellers to follow: Under the physical-presence requirement reaffirmed here, they will not be subjected to use tax collection if they have no physical presence in the taxing State; under the National Geographic rule, mail-order sellers will be subject to use tax collection if they have some presence in the taxing State even if that activity has no relation to the transaction being taxed. See National Geographic, 430 U. S., at 560-562. Between these narrow lines lies the issue of what constitutes the requisite “physical presence” to justify imposition of use tax collection responsibilities.

Instead of confronting this question head on, the majority offers only a cursory analysis of whether Quill’s physical presence in North Dakota was sufficient to justify its use tax collection burdens, despite briefing on this point by the State.3 See Brief for Respondent 45-47. North Dakota contends that even should the Court reaffirm the Bellas Hess rule, Quill’s physical presence in North Dakota was sufficient to justify application of its use tax collection law. Quill concedes it owns software sent to its North Dakota customers, but suggests that such property is insufficient to justify a finding of nexus. In my view, the question of Quill’s actual physical presence is sufficiently close to cast doubt on the majority’s confidence that it is propounding a truly “bright-line” rule. Reasonable minds surely can, and will, differ over what showing is required to make out a “physical pres*331ence” adequate to justify imposing responsibilities for use tax collection. And given the estimated loss in revenue to States of more than $3.2 billion this year alone, see Brief for Respondent 9, it is a sure bet that the vagaries of “physical presence” will be tested to their fullest in our courts.

The majority next explains that its “bright-line” rule encourages “settled expectations” and business investment. Ante, at 316. Though legal certainty promotes business confidence, the mail-order business has grown exponentially despite the long line of our post-Bellas Hess precedents that signaled the demise of the physical-presence requirement. Moreover, the Court’s seeming but inadequate justification of encouraging settled expectations in fact connotes a substantive economic decision to favor out-of-state direct marketers to the detriment of other retailers. By justifying the Bellas Hess rule in terms of “the mail-order industry’s dramatic growth over the last quarter century,” ante, at 316, the Court is effectively imposing its own economic preferences in deciding this case. The Court’s invitation to Congress to legislate in this area signals that its preferences are not immutable, but its approach is different from past instances in which we have deferred to state legislatures when they enacted tax obligations on the States’ shares of interstate commerce. See, e. g., Goldberg v. Sweet, 488 U. S. 252 (1989); Commonwealth Edison Co. v. Montana, 453 U. S. 609 (1981).

Finally, the Court accords far greater weight to stare deci-sis than was given to that principle in Complete Auto itself. As that case demonstrates, we have not been averse to overruling our precedents under the Commerce Clause when they have become anachronistic in light of later decisions. See Complete Auto, 430 U. S., at 288-289. One typically invoked rationale for stare decisis — an unwillingness to upset settled expectations — is particularly weak in this case. It is unreasonable for companies such as Quill to invoke a “settled expectation” in conducting affairs without being taxed. Neither Quill nor any of its amici point to any investment deci*332sions or reliance interests that suggest any unfairness in overturning Bellas Hess. And the costs of compliance with the rule, in light of today’s modern computer and software technology, appear to be nominal. See Brief for Respondent 40; Brief for State of New Jersey as Amicus Curiae 18. To the extent Quill developed any reliance on the old rule, I would submit that its reliance was unreasonable because of its failure to comply with the law as enacted by the North Dakota State Legislature. Instead of rewarding companies for ignoring the studied judgments of duly elected officials, we should insist that the appropriate way to challenge a tax as unconstitutional is to pay it (or in this case collect it and remit it or place it in escrow) and then sue for declaratory judgment and refund.4 Quill’s refusal to comply with a state tax statute prior to its being held unconstitutional hardly merits a determination that its reliance interests were reasonable.

The Court hints, but does not state directly, that a basis for its invocation of stare decisis is a fear that overturning Bellas Hess will lead to the imposition of retroactive liability. Ante, at 317, 318, n. 10. See James B. Beam Distilling Co. v. Georgia, 501 U. S. 529 (1991). As I thought in that case, such fears are groundless because no one can “sensibly insist on automatic retroactivity for any and all judicial decisions in the federal system.” Id., at 546 (White, J., concurring in judgment). Since we specifically limited the question on which certiorari was granted in order not to consider the potential retroactive effects of overruling Bellas Hess, I believe we should leave that issue for another day. If indeed fears about retroactivity are driving the Court’s decision in this case, we would be better served, in my view, to address *333those concerns directly rather than permit them to infect our formulation of the applicable substantive rule.

Although Congress can and should address itself to this area of law, we should not adhere to a decision, however right it was at the time, that by reason of later cases and economic reality can no longer be rationally justified. The Commerce Clause aspect of Bellas Hess, along with its due process holding, should be overruled.