delivered the opinion of the Court.

Petitioner Jones is Director of the Department of Weights and Measures in Riverside County, Cal.1 In that capacity he ordered removed from sale bacon packaged by respondent Rath Packing Co. and flour packaged by three millers, respondents General Mills, Inc., Pillsbury Co., and Seaboard Allied Milling Corp. (hereafter millers). Jones acted after determining, by means of procedures set forth in 4 Cal. Admin. Code c. 8, Art. 5, that the packages were contained in lots2 whose average net weight was less than the net weight stated on the packages. The removal orders were authorized by Cal. Bus. & Prof. Code § 12211 (West Supp. 1977).3

*523Rath and the millers responded by filing suits in the District Court for the Central District of California.4 They sought both declarations that § 12211 and Art. 5 are pre*524empted by federal laws regulating net-weight labeling and injunctions prohibiting Jones from enforcing those provisions. The District Court granted the requested relief5 and, insofar as is relevant here, the Court of Appeals affirmed.6 We granted Jones’ petition for certiorari, 425 U. S. 933 (1976),7 and now affirm the judgments of the Court of Appeals.

I

In its present posture, this litigation contains no claim that the Constitution alone denies California power to enact *525the challenged provisions.8 We are required to decide only whether the federal laws which govern respondents’ packing operations preclude California from enforcing § 12211, as implemented by Art. 5.

Our prior decisions have clearly laid out the path we must follow to answer this question. The first inquiry is whether Congress, pursuant to its power to regulate commerce, U. S. Const., Art. 1, § 8, has prohibited state regulation of the particular aspects of commerce involved in this case. Where, as here, the field which Congress is said to have pre-empted has been traditionally occupied by the States, see, e. g., U. S. Const., Art. I, § 10; Patapsco Guano Co. v. North Carolina, 171 U. S. 345, 358 (1898), “we start with the assumption that the historic police powers of the States were not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress.” Rice v. Santa Fe Elevator Corp., 331 U. S. 218, 230 (1947). This assumption provides assurance that “the federal-state balance,” United States v. Bass, 404 U. S. 336, 349 (1971), will not be disturbed unintentionally by Congress or unnecessarily by the courts. But when Congress has “unmistakably . . . ordained,” Florida Lime & Avocado Growers, Inc. v. Paul, 373 U. S. 132, 142 (1963), that its enactments alone are to regulate a part of commerce, state laws regulating that aspect of commerce must fall. This result is compelled whether Congress’ command is explicitly stated in the statute’s language or implicitly contained in its structure and purpose. City of Burbank v. Lockheed Air Terminal, Inc., 411 U. S. 624, 633 (1973); Rice v. Santa Fe Elevator Corp., supra, at 230.

Congressional enactments that do not exclude all state legislation in the same field nevertheless override state laws *526with which they conflict. U. S. Const., Art. VI. The criterion for determining whether state and federal laws are so inconsistent that the state law must give way is firmly established in our decisions. Our task is “to determine whether, under the circumstances of this particular case, [the State’s] law stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” Hines v. Davidowitz, 312 U. S. 52, 67 (1941). Accord, De Canas v. Bica, 424 U. S. 351, 363 (1976); Perez v. Campbell, 402 U. S. 637, 649 (1971); Florida Lime & Avocado Growers, Inc. v. Paul, supra, at 141; id., at 165 (White, J., dissenting). This inquiry requires us to consider the relationship between state and federal laws as they are interpreted and applied, not merely as they are written. See De Canas v. Bica, supra, at 363-365; Swift & Co. v. Wickham, 230 F. Supp. 398, 408 (SDNY 1964), appeal dismissed, 382 U. S. 111 (1965), aff'd on further consideration, 364 F. 2d 241 (CA2 1966), cert. denied, 385 U. S. 1036 (1967).

II

Section 12211 of the Cal. Bus. & Prof. Code (West Supp. 1977) applies to both Rath’s bacon and the millers’ flour. The standard it establishes is straightforward: “[T]he average weight or measure of the packages or containers in a lot of any . . . commodity sampled shall not be less, at the time of sale or offer for sale, than the net weight or measure stated upon the package.”

In order to determine whether that standard has been violated, local officials such as Jones follow the statistical sampling procedure set forth in Art. 5.9 That procedure requires the inspector to identify a lot of identical packages of a commodity and determine the number of packages in that lot. *527He then determines, from tables in the regulation, the number of packages necessary to provide a suitable sample of the lot, and a smaller number of packages which is used to determine the average tare.10 After determining that average, the inspector weighs each package in the sample, subtracts the average tare, and records the difference between the measured and the stated net weights. These measurements are used to identify individual packages in the sample which deviate unreasonably from the stated weight. Those packages are replaced11 in the sample and the replacements weighed. *528Finally, the deviations from the stated weight are totaled algebraically and compared with tables which indicate the magnitude of the total error necessary to conclude that the lot’s average weight is or is not less than the stated weight.12

III

A. Rath’s bacon is produced at plants subject to federal inspection under the Federal Meat Inspection Act (FMIA or Act), as amended by the Wholesome Meat Act, 81 Stat. 584, 21 U. S. C. § 601 et seq. Among the requirements imposed on federally inspected plants, and enforced by Department of Agriculture inspectors,13 are standards of accuracy in labeling. On the record before us, we may assume that Rath’s bacon complies with these standards.14

The federal labeling requirement is imposed by § 7 (b) of the FMIA, 81 Stat. 588, 21 U. S. C. § 607 (b), which commands:

“All . . . meat and meat food products inspected at any *529establishment under the authority of this title . . . shall at the time they leave the establishment bear . . . the information required under paragraph (n) of section 1 of this Act.”

Section 1 (n) of the FMIA, 21 U. S. C. § 601 (n), defines the term “misbranded.” As relevant here, it provides that meat or a meat product is misbranded

“(5) if in a package or other container unless it bears a label showing ... (B) an accurate statement of the quantity of the contents in terms of weight, measure, or numerical count: Provided, That . . . reasonable variations may be permitted, and exemptions as to small packages may be established, by regulations prescribed by the Secretary.” 81 Stat. 586.

Other sections of the FMIA prohibit dealing in misbranded products, as defined by § 1 (n).16

The Secretary of Agriculture has used his discretionary authority to permit “reasonable variations” in the accuracy of the required statement of quantity:

“The statement [of net quantity of contents] as it is shown on a label shall not be false or misleading and shall express an accurate statement of the quantity of contents of the container exclusive of wrappers and packing substances. Reasonable variations caused by loss or gain of moisture during the course of good distribution practices or by unavoidable deviations in good manufacturing practice will be recognized. Variations from stated quantity of contents shall not be unreasonably large.” 9 CFR § 317.2 (h)(2) (1976).

Thus, the FMIA, as implemented by statutorily authorized regulations, requires the label of a meat product accurately to indicate the net weight of the contents unless the difference *530between stated and actual weights is reasonable and results from the specified causes.15

B. Section 408 of the FMIA, 21 U. S. C. § 678, prohibits the imposition of “[m]arking, labeling, packaging, or ingredient requirements in addition to, or different than, those made under” the Act.17 This explicit pre-emption provision *531dictates the result in the controversy between Jones and Rath. California’s use of a statistical sampling process to determine the average net weight of a lot implicitly allows for variations from stated weight caused by unavoidable deviations in the manufacturing process.18 But California makes no allowance for loss of weight resulting from moisture loss during the course of good distribution practice.19 Thus, the state *532law’s requirement—that the label accurately state the net weight, with implicit allowance only for reasonable manufacturing variations—is “different than” the federal requirement, which permits manufacturing deviations and variations caused by moisture loss during good distribution practice.

Petitioner Jones seeks to avoid this result by arguing that, the FMIA’s provisions governing the accuracy of the required net-quantity statements are not “labeling requirements” within the meaning of § 408. He contends that “labeling” refers only to the format and placement of information, not to its content.20 Requirements relating to accuracy, according to Jones, deal with the problem of misbranding, and § 408 grants the States concurrent jurisdiction over that subject.

We agree with the Court of Appeals that this argument is “strained.” 530 F. 2d, at 1314 n. 25. Nothing in the Act suggests the restrictive meaning petitioner ascribes to the phrase “labeling requirements.” To the contrary, § 7 (b) requires that the product bear specified information, see supra, at 528-529, and § 1 (p) of the FMIA, 21 U. S. C. § 601 (p),21 makes clear that any material bearing that information is part of the product’s labeling. It twists the language beyond the breaking point to say that a law mandating that labeling contain certain information is not a “labeling requirement.”

We therefore conclude that with respect to Rath’s packaged bacon, § 12211 and Art. 5 are pre-empted by federal law.

IV

A. The federal law governing net-weight labeling of the millers’ flour is contained in two statutes, the Federal Food, Drug, and Cosmetic Act (FDCA), 52 Stat. 1040, as amended, *53321 U. S. C. § 301 et seq., and the Fair Packaging and Labeling Act (FPLA), 80 Stat. 1296, as amended, 15 U. S. C. §§ 1451—1461. For the reasons stated below, we conclude that the federal weight-labeling standard for flour is the same as that for meat.

The FDCA prohibits the introduction or delivery for introduction into interstate commerce of any food22 that is misbranded. 21 U. S. C. § 331. A food is misbranded under the FDCA,

“[i]f in package form unless it bears a label containing . . . an accurate statement of the quantity of the contents in terms of weight, measure, or numerical count: Provided, That . . . reasonable variations shall be permitted, and exemptions as to small packages shall be established, by regulations prescribed by the Secretary.” § 343 (e).

This provision is identical to the parallel provision in the FMIA, see supra, at 529, except that the FDCA mandates rather than allows the promulgation of implementing regulations.23 The regulation issued in response to this statutory mandate is also substantially identical to its counterpart under the FMIA:

“The declaration of net quantity of contents shall express an accurate statement of the quantity of contents of the package. Reasonable variations caused by loss or gain of moisture during the course of good distribution practice or by unavoidable deviations in good manufacturing practice will be recognized. Variations from stated quantity of contents shall not be unreasonably large.” 21 CFR § 1.8b (q) (1976).

*534Since flour is a food under the FDCA, its manufacture is also subject to the provisions of the FPLA. See 15 U. S. C. §§ 1452, 1459 (a). That statute states a congressional policy that “[p]ackages and their labels should enable consumers to obtain accurate information as to the quantity of the contents and should facilitate value comparisons.” § 1451. To accomplish those goals, insofar as is relevant here, the FPLA bans the distribution in commerce of any packaged commodity unless it complies with regulations

“which shall provide that—

“(2) The net quantity of contents (in terms of weight, measure, or numerical count) shall be separately and accurately stated in a uniform location upon the principal display panel of [the required] label.” § 1453 (a).

The FPLA also contains a saving clause which specifies that nothing in the FPLA “shall be construed to repeal, invalidate, or supersede” the FDCA. § 1460. Nothing in the FPLA explicitly permits any variation between stated weight and actual weight.

The amici States contend that since the FPLA does not allow any variations from stated weight, there is no difference between federal law governing labeling of flour and California law. The Court of Appeals, however, held that because of the saving clause, compliance with the FDCA, which does allow reasonable variations, satisfies the requirements of the FPLA. 530 F. 2d, at 1325. Amici respond that the Court of Appeals misinterpreted the FDCA and that the FDCA establishes a statutory standard of strict accuracy for net-weight labeling. They argue, therefore, that the saving clause of the FPLA does not alter the standard mandated by § 1453. Brief for 39 States as Amici Curiae 15-21. Alternatively, the States argue that although the saving clause means that the FPLA does not supersede the FDCA, “it *535cannot be construed to excuse compliance with FPLA standards where both FDCA and FPLA requirements are applicable.” Id., at 28.

The States’ argument that the FDCA standard makes no allowance for reasonable variations is based on this Court’s opinion in United States v. Shreveport Grain & Elevator Co., 287 U. S. 77 (1932). Shreveport decided an appeal by the Government in a criminal case involving shortweighting in violation of the predecessor of the FDCA, the Food and Drugs Act, 34 Stat. 768, as amended, c. 117, 37 Stat. 732. The trial court had dismissed the indictment under that statute, which was essentially identical to the net-weight labeling requirement of the FDCA,24 on the ground that the prohibition of unreasonable variations from the marked weight was too indefinite to state a criminal offense. We reversed, holding that the statute’s substantive standard was created by the “accurate statement” language which preceded the proviso allowing reasonable variations, and that the proviso merely granted administrative authority to promulgate regulations permitting variations “from the hard and fast rule of the act.” 287 U. S., at 81-82. Since Congress re-enacted the language interpreted by the Shreveport Court, FDCA, c. 675, § 403 (e), 52 Stat. 1047, amici conclude that the standard under the FDCA is also a “hard and fast rule.”

We need not decide whether the rationale as well as the *536result of Shreveport remains good law.25 It is clear that 21 CFR § 1.8b (q) (1976), insofar as it is based on the FDCA, has the force of law26 and allows reasonable variations. Thus, whether the statutory standard is viewed as strict, with the regulation considered a restriction on the power to prosecute, or whether the standard is itself viewed as incorporating the flexibility of the proviso and its implementing regulation,27 the result is the same. Under the FDCA, reasonable variations from the stated net weight do not subject a miller to prosecution, whether civil or criminal, if the variations arise from the permitted causes. The question raised by the arguments of amici is whether by enacting the FPLA, Congress intended to eliminate the area of freedom from prosecution created by the FDCA and its implementing regulation.

Over 60 years ago, Congress concluded that variations must be allowed because of the nature of certain foods and the impossibility of developing completely accurate means of packing. H. R. Rep. No. 850, 62d Cong., 2d Sess., 2 *537(1912); S. Rep. No. 1216, 62d Cong., 3d Sess., 2-3 (1913).28 Since 1914, regulations under the food and drug laws have permitted reasonable variations from stated net weight resulting from packing deviations or gain or loss of moisture occurring despite good commercial practice. See United States v. Shreveport Grain & Elevator Co., supra, at 84. If Congress had intended to overrule this longstanding administrative practice, founded on a legislative statement of necessity, we would expect it to have done so clearly. Instead, it explicitly preserved existing law, with “no changes.” 15 U. S. C. § 1460; S. Rep. No. 1186, 89th Cong., 2d Sess., 20 (1966). The legislative history of the FPLA contains some indication that the saving clause was understood to preserve the reasonable-variation regulation under the FDCA,29 and no evidence that Congress affirmatively intended to overrule that regulation.30 We can only conclude that under the FPLA, as under the FDCA, a manufacturer of food is not *538subject to enforcement action for violation of the net-weight labeling requirements if the label accurately states the net weight, with allowance for the specified reasonable variations.

B. The FDCA contains no pre-emptive language. The FPLA, on the other hand, declares that

“it is the express intent of Congress to supersede any and all laws of the States or political subdivisions thereof insofar as they may now or hereafter provide for the labeling of the net qua[nt]ity of contents of the package of any consumer commodity covered by this chapter which are less stringent than or require information different from the requirements of section 1453 of this title or regulations promulgated pursuant thereto.” 15 U. S. C. § 1461.31

The Court of Appeals, although recognizing that this section leaves more scope for state law than does the FMIA, concluded that § 12211, as implemented by Art. 5, is pre-empted because it is less stringent than the Federal Acts, 530 F. 2d, at 1324-1327.

The basis for the Court of Appeals’ holding is unclear. Its opinion may be read as based on the conclusion that the state law is inadequate because its enforcement relies on a statistical averaging procedure. We have rejected that conclusion. See supra, at 531, and n. 18. Alternatively, the Court of Appeals may have found California’s approach less stringent because the State takes no enforcement action against lots whose *539average net weight exceeds the weight stated on the label, even if that excess is not a reasonable variation attributable to a federally allowed cause.

We have some doubt that by pre-empting less stringent state laws, Congress intended to compel the States to expend scarce enforcement resources to prevent the sale of packages which contain more than the stated net weight. We do not have to reach that question, however, because in this respect California law apparently differs not at all from federal law, as applied. The inspectors responsible for enforcing the net-weight labeling provisions of the Federal Acts are officially informed that “[f]ield weighing for net weight is primarily to determine the likelihood of short weight units in the lots.” Moreover, they are not required to submit samples to headquarters “if the average net is not below the amount declared on the label.” Food and Drug Administration, Inspection Operations Manual 448.1, 448.13 (1976). These instructions undercut the argument that there is a federal interest in preventing packages from being overfilled.32 Since neither jurisdiction is concerned with overweighting in the administration of its weights and measures laws, we cannot say that California’s statutory lack of concern for that “problem” 33 makes its laws less stringent than the federal.

*540Respondents argue that California’s law is pre-empted because it requires information different from that required by federal law. The meaning of the statutory pre-emption of laws that require “information different from” the federal net-weight labeling provisions, like the meaning of the phrase “less stringent,” is unclear. Respondents attribute to the ban on requiring different information a broad meaning, similar in scope to the pre-emption provision of the FMIA. They contend that since California law requires the label to state the minimum net weight, it requires “information different from” the federal laws, which demand an accurate statement with allowance for the specified reasonable variations. Brief for Respondents 31-32. The legislative history, however, suggests that the statute expressly pre-empts as requiring “different information” only state laws governing net quantity labeling which impose requirements inconsistent with those imposed by federal law.34 Since it would be possible to comply with the state law without triggering federal enforcement action we conclude that the state requirement is not inconsistent with federal law. We therefore hold that 15 U. S. C. § 1461 does not pre-empt California’s § 12211 as implemented by Art. 5.

That holding does not, however, resolve this case, for we still must determine whether the state law “stands as an obstacle to the accomplishment and execution of the full *541purposes and objectives of Congress.” See supra, at 526. As Congress clearly stated, a major purpose of the FPLA is to facilitate value comparisons among similar products. Obviously, this goal cannot be accomplished unless packages that bear the same indicated weight in fact contain the same quantity of the product for which the consumer is paying. The significance of this requirement for our purposes results from the physical attributes of flour.

Flour is composed of flour solids and moisture. The average water content of wheat kernels used to make flour is 12.5% by weight, with a range from 10% to 14.5%. Efficient milling practice requires adding water to raise the moisture content to 15% to 16%; if the wheat is too wet or too dry, milling will be hindered. During milling, the moisture content is reduced to 13% to 14%. App. 28-29.35

The moisture content of flour does not remain constant after milling is completed. If the relative humidity of the atmosphere in which it is stored is greater than 60%, flour will gain moisture, and if the humidity is less than 60%, it will lose moisture.36 The federal net-weight labeling standard permits variations from stated weight caused by this gain or loss of moisture.

Packages that meet the federal labeling requirements37 *542and that have the same stated quantity of contents can be expected to contain the same amount of flour solids.38 Manufacturers will produce flour with a moisture content fixed by the requirements of the milling process.39 Since manufacturers have reason not to pack significantly more than is required and federal law prohibits underpacking, they will pack the same amount of this similarly composed flour into packages of any given size.40 Despite any changes in weight resulting from changes in moisture content during distribution, the packages will contain the same amount of flour solids when they reach the consumer. This identity of contents facilitates consumer value comparisons.

The State’s refusal to permit reasonable weight variations resulting from loss of moisture during distribution produces a different effect.41 In order to be certain of meeting the California standard, a miller must ensure that loss of moisture during distribution will not bring the weight of the contents below the stated weight. Local millers, which serve a limited area, could do so by adjusting their packing practices to the specific humidity conditions of their region. For example, a miller in an area where the humidity is typically higher than *54360% would not need to overpack at all. By contrast, a miller with a national marketing area would not know the destination of its flour when it was packaged and would therefore have to assume that the flour would lose weight during distribution. The national manufacturer, therefore, would have to overpack.

Similarly, manufacturers who distributed only in States that followed the federal standard would not be concerned with compensating for possible moisture loss during distribution. National manufacturers who did not exclude the nonconforming States from their marketing area, on the other hand, would have to overpack. Thus, as a result of the application of the California standard, consumers throughout the country who attempted to compare the value of identically labeled packages of flour would not be comparing packages which contained identical amounts of flour solids. Value comparisons which did not account for this difference—and there would be no way for the consumer to make the necessary calculations—would be misleading.

We therefore conclude that with respect to the millers’ flour, enforcement of § 12211, as implemented by Art. 5, would prevent “the accomplishment and execution of the full purposes and objectives of Congress” in passing the FPLA. Under the Constitution, that result is impermissible, and the state law must yield to the federal.

The judgments are affirmed.

It is so ordered.

with whom Mr. Justice Stewart joins, concurring in part and dissenting in part.

I agree that with respect to Rath’s packaged bacon, § 12211 of the Cal. Bus. & Prof. Code and Art. 5 of 4 Cal. Admin. Code, c. 8, are pre-empted by the express pre-emptive provision of the Federal Meat Inspection Act, 21 U. S. C. § 678. I also agree that with respect to General Mills’ flour, § 12211 *544and Art. 5 are not pre-empted by the express pre-emptive provision of the Fair Packaging and Labeling Act (FPLA), 15 U. S. C. § 1461. I am unable to agree, however, with the implicit pre-emption the Court finds with respect to the flour. This latter pre-emption is founded in unwarranted speculations that hardly rise to that clear demonstration of conflict that must exist before the mere existence of a federal law may be said to pre-empt state law operating in the same field.

With respect to labeling requirements for flour under the scheme contemplated by the FPLA in conjunction with the Federal Food, Drug, and Cosmetic Act, the Court determines that the state-law labeling requirements are neither “less stringent than” nor inconsistent with those federal requirements. This conclusion quite properly dictates the Court’s holding that Congress has not expressly prohibited state regulation in this field. The remaining inquiry, then, is whether the two statutory schemes are in utter conflict.1 As this Court noted in Kelly v. Washington, 302 U. S. 1, 10 (1937):

“The principle is thoroughly established that the exercise by the State of its police power, which would be valid if not superseded by federal action, is superseded only where the repugnance or conflict is so 'direct and positive’ that the two acts cannot 'be reconciled or consistently stand together.’ ”

See also Cloverleaf Butter Co. v. Patterson, 315 U. S. 148, 156 (1942); Askew v. American Waterways Operators, Inc., 411 U. S. 325, 337, 341 (1973). When we deal, as we do here, with congressional action “in a field which the States have traditionally occupied,” the basic assumption from which pre-*545emption must be viewed is “that the historic police powers of the States were not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress.” Rice v. Santa Fe Elevator Corp., 331 U. S. 218, 230 (1947); cf. De Canas v. Bica, 424 U. S. 351, 356 (1976). I am simply unable to find that this stringent standard has been met in this case.

The Court’s opinion demonstrates that it is physically possible to comply with the state-law requirement “without triggering federal enforcement action,” ante, at 540. This leads the Court to conclude that the “state requirement is not inconsistent with federal law.” Ibid. It also must lead to the conclusion that this is not a case “where compliance with both federal and state regulations is a physical impossibility for one engaged in interstate commerce.” Florida Lime & Avocado Growers, Inc. v. Paul, 373 U. S. 132, 142-143 (1963). Pre-emption, then, if it is to exist at all in this case, must exist because the operation of the state Act inexorably conflicts with the purposes underlying the Federal Act. The Court relies on the fact that one of the purposes of the FPLA is to “facilitate value comparisons” among consumers, 15 U. S. C. § 1451. But merely identifying a purpose is not enough; it must also be shown that the state law inevitably frustrates that purpose. As we but recently noted:

“We must also be careful to distinguish those situations in which the concurrent exercise of a power by the Federal Government and the States or by the States alone may possibly lead to conflicts and those situations where conflicts will necessarily arise. 'It is not . . . a mere possibility of inconvenience in the exercise of powers, but an immediate constitutional repugnancy that can by implication alienate and extinguish a pre-existing right of [state] sovereignty.’ The Federalist No. 32, p. 243 (B. Wright ed. 1961).” Goldstein v. California, 412 U. S. 546, 554-555 (1973) (emphasis in original).

*546Under the proper test, it is only

“[i]f the purpose of the act cannot otherwise be accomplished—if its operation within its chosen field else must be frustrated and its provisions be refused their natural effect—the state law must yield to the regulation of Congress within the sphere of its delegated power.” Savage v. Jones, 225 U. S. 501, 533 (1912).

The Court's reliance on supposition and inference fails in two respects to demonstrate that respondents have carried their burden of demonstrating pre-emption. First, on the Court’s own premises, there should be no finding of pre-emption. We are told, ante, at 526, that the relevant inquiry is “the relationship between state and federal laws as they are interpreted and applied, not merely as they are written,” while we are further told, ante, at 539, that there is, in fact, no “federal interest in preventing packages from being overfilled,” since the Federal Government is not “concerned with overweighting in the administration of its weights and measures laws . . . .” Under these premises, it is hard to accept the Court’s conclusion that, because of the federal purpose to facilitate consumer value comparisons,2 the state law is pre-empted because some packages might contain more than the minimum weight stated and more than another company’s similarly marked package. For, we have been told that, should a manufacturer deliberately overpack, for whatever reason,3 there will be no federal action taken against him even though value comparisons might then “be misleading.” It is virtually impossible to say, as the Court does, that “neither the State nor the Federal Government is concerned with over*547weighting,” ante, at 542 n. 41, and yet conclude that state-induced overweighting conflicts with a “value comparison” purpose, while, presumably, other overweighting does not. In viewing such a purpose to be sufficient to require pre-emption while the very purpose is ignored in practice by the administering federal agency reverses the normal presumption against finding pre-emption. The reasoning process which leads the Court to conclude that there is no express pre-emption, ante, at 540, leads me to conclude that there is no implied pre-emption.4

Second, and as troubling as the legal inconsistency, is the Court’s reliance on unproved factual speculation in demonstrating the purported irreconcilable undermining of the federal purpose by the state statutory scheme. The premises the opinion must rely on are many. It acknowledges that flour packed under different humidity conditions would nonetheless comply with the federal standard, even though, as a result, similarly marked packages might contain different quantities of flour “solids,” ante, at 542, and n. 39, but relies on the economics of the milling process to conclude that packers “will pack the same amount of [flour solids] into packages of any given size.” This may normally be true as an economic fact, but it is not supported by the record and as a Court we have no way of knowing it from other sources.

Similarly defective is the reasoning process by which the majority concludes that local millers could adjust their packaging practices to specific humidity conditions, while national millers could not, since the national millers “would not know the destination of [their] flour when it was packaged and would therefore have to assume that the flour would lose weight during distribution.” Ante, at 543. This assump*548tion, too, is unsupported by the record.5 We simply have no basis for concluding that national distributors do not know, or could not know through the exertion of some modicum of effort, where their flour will end up. The possibility that a packer might have to incur some extra expense in meeting both systems simply does not mean that the “purposes of the act cannot otherwise be accomplished,” Savage v. Jones, 225 U. S., at 533, nor does it demonstrate that “the two acts cannot 'be reconciled . . . ,' ” Kelly v. Washington, 302 U. S., at 10.6

*549The assumptions in the Court’s opinion not only are insufficient to compel a finding of implied pre-emption, they suggest an approach to the question of pre-emption wholly at odds with that enunciated in Florida Lime & Avocado Growers, Inc. v. Paul, 373 U. S. 132 (1963). There, this Court was concerned with differing federal and state maturity standards for avocados grown in Florida. This Court rejected a test which looked to the similarity of purposes, id., at 142, and noted instead that a manufacturer could have complied with both statutes by modifying procedures somewhat, id., at 143, which demonstrated that there was “no inevitable collision between the two schemes of regulation, despite the dissimilarity of the standards,” ibid. Nothing has been shown to demonstrate that this conclusion is not equally justified in the instant case.

The Court today demonstrates only that there could be— not that there must be—a conflict between state and federal laws.7 Because reliance on this test to find pre-emption, absent an explicit pre-emptive clause, seriously misapprehends the carefully delimited nature of the doctrine of pre-emption, Goldstein v. California, 412 U. S., at 554, I dissent from the holding that § 12211 and Art. 5 are pre-empted with respect to General Mills’ flour.