with whom Justice Stevens joins, concurring in part and dissenting in part.
The question before us is the proper construction of the access-to-records clauses required by law to be inserted in various Government contracts. These clauses authorize the Comptroller General to inspect any records that “directly pertain to and involve transactions relating to” the contract into which the clause was inserted. 10 U. S. C. § 2313(b); see 41 U. S. C. § 254(c) (permitting access to “directly pertinent” records “involving transactions related to” the contracts). The Court now holds that these clauses permit access to certain cost records, even when the contract is not cost-based and was negotiated without regard to costs. I cannot agree that cost records are “directly pertinent” to non-cost-based contracts. I therefore dissent insofar as the Court permits access to such records.
The Court correctly begins its analysis by focusing on the language of the statute, noting that it contains “words of limitation designed to restrict the class of records to which access is permitted by requiring some close connection between the type of records sought and the particular contract.” Ante, at 831. But after this nod in the direction of the statutory language, the Court, unaccountably in my estimation, fails to examine the type of records sought, the nature of the particular contract, and the closeness of the connection between them. Instead, the Court wanders off to explore the general “aims embraced by Congress in enacting the access-to-records provisions”: protecting contractors from govern*861mental intrusion, and equipping the General Accounting Office with a tool to detect fraud and waste in Government contracting. Ante, at 833. Given the “tension” between these “dual, conflicting aims,” the Court concludes that it must “balance the public interest served by full GAO investigations against the private interest in freedom from officious governmental intermeddling in the contractor’s private business affairs.” Ante, at 834-835. The Court then reaches what it deems an “appropriate accommodation of the competing goals reflected in the legislative history” by holding that the Comptroller General may inspect records of “direct costs,” but is barred from inspecting other records except to the extent they are allocated to the particular contract. Ante, at 839-840.
I agree that the legislative history reflects conflicting congressional goals, but it appears to me that the necessary “balancing” and “accommodation” of these goals has already been accomplished by Congress. Congress may well have intended to give the GAO a tool to detect fraud and waste and to improve the procurement process, but the particular tool it crafted was a limited one. It gave the Comptroller General access to a narrow category of records: those directly pertinent to the contracts between the Government and its contracting partner.
The Court asserts that Merck’s records of direct costs are directly pertinent to its contracts with the Government because such costs “are, by definition, readily identifiable as attributable to the specific product supplied under the contract.” Ante, at 840 (emphasis added). Records of indirect costs are not pertinent to Merck’s contracts, the Court says, because such records are “completely unrelated to either the contract underlying the access demand or the product procured under that contract.” Ante, at 836 (emphasis added). But, as the Court obviously recognizes, there is a difference between records pertinent to the “contracts” by which the Government does its purchasing, and records pertinent to *862“products” ultimately purchased under those contracts.1 In interpreting the access-to-records statutes we must “presume that Congress chose its words with . . . care,” FBI v. Abramson, 456 U. S. 615, 635 (1982) (O’Connor, J., dissenting), and Congress chose to require pertinence to the specific “contract” at issue, not to the “product” procured.
It is a fundamental canon of statutory construction that “unless otherwise defined, words will be interpreted as taking their ordinary, contemporary, common meaning.” Perrin v. United States, 444 U. S. 37, 42 (1979). In its ordinary meaning, a “contract” is a legally enforceable bargain, formed by mutual consent and supported by consideration. Restatement (Second) of Contracts §17 (1979); 1 W. Jaeger, Williston on Contracts § 18 (3d ed. 1957). Particular goods and services may be the subject matter of a contract, but the “contract” itself is the agreement between the parties. Nothing in the legislative history suggests that Congress in-' tended any other meaning by its choice of the word “contract.”2 The access-to-records statutes were intended to *863permit inspection of cost records when the negotiation of a particular contract depended on representations about the contractor’s costs, or when the contract price itself varied according to costs. See 97 Cong. Rec. 13198 (1951) (remarks of Rep. Hardy).3 There is, however, no indication that Con*864gress intended to permit such inspection when costs were not relevant to the contract’s negotiation, its terms, or its performance.
Apparently recognizing that cost records are not accessible to the Comptroller General unless they bear some relationship to the actual contracts at issue, the Court asserts that direct cost records are pertinent to non-cost-based contracts because “the contractor will have some regard for these costs in setting even a catalog price in order to avoid a loss on the product.” The Court concludes that “these costs therefore have a very direct influence on the price charged the Government.” Ante, at 840.4 But the fact that costs may affect a seller’s decision to offer a certain price does not make costs directly pertinent to the contract. A contract, after all, is a meeting of the minds. Many factors may affect one party’s willingness to make an offer or the other party’s willingness to accept it, but the vast majority of these factors are not mentioned in the bargaining process and play no part in the agreement ultimately reached. An unarticulated motive for one party’s assent to a bargain is not pertinent, much less “directly pertinent,” to the bargain itself.
Pursuant to the contracts at issue here, Merck has promised to provide various pharmaceutical products and the Government has promised to pay a certain price. The Government’s promise was based on Merck’s standard catalog price for the products purchased. Information about Merck’s pro*865duction costs was not requested, and neither the contract price nor its performance was tied to costs in any respect. Merck’s costs were thus irrelevant to the bargain reached by the parties. Because the Government chose not to make costs an issue during the negotiations, the terms of the contracts would have been the same whether Merck’s costs represented 1 percent, 10 percent, or 100 percent of the price the Government agreed to pay. The conclusion seems inescapable that Merck’s costs, and consequently its cost records, are not pertinent to the agreements it entered into with the Government.5
If the Government is concerned that it may be paying too high a price for pharmaceutical products, it is always free to ask for cost data in connection with any contract it negotiates in the future. Cost records then would be pertinent, because representations about costs would form part of the basis for the bargain between the parties. In this case, however, costs were not relevant to the Government’s decision to contract with Merck. Cost records do not become “directly pertinent” to these contracts simply because the Government now wants information that did not concern it at the time of contracting.
I would hold that when the terms of a contract are not tied to costs and the contractor makes no representations about costs during its negotiations with the Government, cost records do not “directly pertain to and involve transactions relating to the contract” and are not subject to inspection by the Comptroller General. The only records “directly pertinent” to such a contract would be those necessary to verify the terms of the contract and the representations upon which *866the contract was based. In this case, the Comptroller General’s access would be limited to those records necessary to verify that Merck actually had an established catalog price for the products it sold, that it sold the products in substantial quantities to the general public at the catalog price, that it delivered the product specified, and that it received from the Government no more than the amount due under the contract. To the extent the Court concludes that the Comptroller General may obtain Merck’s cost records as well, I dissent.
For example, Merck’s research notes, pharmaceutical formulas, and other trade secrets might be records directly pertinent to the “products” purchased by the Government, and might be of assistance to the Government in evaluating the success of its procurement policies. But unless these records were also pertinent to Merck’s contracts with the Government, they would not be available to the Comptroller General under his access-to-records authority.
The first Court of Appeals case interpreting the access-to-records statutes, Hewlett-Packard Co. v. United States, 385 F. 2d 1013 (CA9 1967), cert. denied, 390 U. S. 988 (1968), reached a result similar to that reached by the Court today. In Hewlett-Packard, the Court of Appeals recognized that if the word “contract” were given its ordinary meaning, to refer to “the terms and conditions of an agreement and the procedures whereby those terms and conditions were formulated,” cost records could not be regarded as “directly pertinent” to a non-cost-based contract. 385 F. 2d, at 1016. In an effort to provide the Government with greater access, however, the Court of Appeals invented its own definition of the word “contract.” Without explanation or citation of authority, the Court of Appeals declared:
*863“[T]he word ‘contract,’ as used in this statute is intended to have a broader meaning, embracing not only the specific terms and conditions of the agreement, but also the general subject matter. The subject matter of these four contracts is the procurement of described property by the Government.” Ibid.
The court concluded that cost records “directly pertain to that subject matter, because if out of line with the contract price, the contract may have been an inappropriate means of meeting this particular procurement need of the Government.” Ibid.
Although the Court in the present case does not explicitly adopt this strained reading of the word “contract,” it approves the so-called “Bristol test” which was developed partly in reliance on Hewlett-Packard, and partly in response to a decision by one of the parties voluntarily to surrender certain types of records. See Bristol Laboratories Division of Bristol-Myers Co. v. Staats, 428 F. Supp. 1388 (SDNY 1977), aff’d, 620 F. 2d 17 (CA2 1980), aff’d by an equally divided Court, 451 U. S. 400 (1981). The Court of Appeals in this case and the Court of Appeals for the Third Circuit, SmithKline Corp. v. Staats, 668 F. 2d 201 (1981), cert. pending, Nos. 81-2082, 81-2268, have applied the “Bristol test” without independent analysis. In adopting this test, the Court never explains whether it agrees with the definition of “contract” developed in Hewlett-Packard, or whether it believes some other definition leads to the same result.
Representative Hardy, the sponsor of the legislation, gave several examples of contracts in which the access-to-records clauses would be useful. The first was a construction contract in which the Government agreed to pay a fixed overhead rate “as a proportion of direct labor costs or as a lump sum . . . based upon the experience of the contractor in performing similar work in the past.” The second was a pair of contracts, one cost-plus and one fixed-price, in which “the contractor was charging to the [cost-plus] contract expenses which should have been charged to the fixed-price contract.” The third' was a contract with a price redetermination clause, authorizing “an increase or decrease in the price payable under the contract as the work progresses or after it has been completed.” 97 Cong. Rec. 13198 (1951). In each of these three examples, the contractor’s costs were highly relevant to the negotiation or performance of the contract, and cost records thus would have been pertinent to the contract. But neither Rep*864resentative Hardy nor any other Member of Congress suggested that cost records would be pertinent to a contract in which the price was negotiated without regard to the contractor’s costs.
When a contract is not cost-based, the relationship between direct costs and prices may be tenuous. In these cases, there is every reason to suppose that Merck’s prices are not related to its direct costs; the Court notes, ante, at 842, n. 20, the Government’s suggestion that “direct costs may represent as little as 9% of the sales price of a pharmaceutical product.” Merck may have taken direct costs into account in setting its catalog price, but it may equally well have relied on other factors, such as market surveys or prices set by its competitors.
1 agree with the Court that the Hoffman amendment, adding the word “directly” before “pertinent” in the statute, was intended to “circumscribe the inquiry that the Comptroller General was authorized to undertake.” Ante, at 833. I find it unnecessary to place much emphasis on this amendment, however, because I do not regard cost records as being at all pertinent to a non-cost-based contract.