delivered the opinion of the Court.

El Paso Natural Gas Co. is an interstate natural gas pipeline company that delivers gas at the Arizona-California border to three California distribution companies. The present controversy concerns gas to be purchased by it in Texas from Lo-Vaca Gathering Co. and Houston Pipe Line Co. Under Lo-Vaca’& contract gas produced in Texas is to be delivered to a subsidiary of El Paso’S at a Texas point for delivery, into its pipeline. The contract contains the following two clauses:

“All of the gas to be purchased by El Paso from Gatherer [Lo-Vaca] under this agreement shall be used by El Paso solely as fuel in El Paso’s compressors, treating plants, boilers, camps and other facilities locátéd outside of the State of Texas. It is understood, however, that said gas will be commingled with other gas being transported in El Paso’s pipe fine system.”

“It is the intent and understanding of the parties hereto that the sale of natural gas hereof is not subject to the jurisdiction of the Federal Power Commission because this sale is not for resale.”

This “restricted use” agreement provides for a Separate metering of the contract volumes prior to their delivery info El Paso’s system. El Paso will meter the. gas used *368for fuel purposes in its New Mexico and Arizona facilities to make certain this amount invariably exceeds the volumes of gas taken from Lo-Vaca under this agreement.

El Paso and Houston made a similar contract containing a similar “restricted use” provision by which El Paso covenants that this Houston gas will be consumed by El Paso solely as fuel in its Texas operations or in another Texas plant. This contract, like the other one, also provides for metering the volume of gas delivered in Texas; and it includes a covenant by El Paso that the Texas uses will &t all times exceed the amounts supplied by Houston.

In spite of these "restricted use” covenants it is conceded that the gas sold by Lo-Vaca and Houston to El Paso will flow in a commingled stream with gas from other sources and that at least a portion of the gas will in fact be resold out of Texas.

The Federal Power Commission asserted jurisdiction over these sales as sales in interstate commerce “for resale,” as that term is used in § 1 (b) of the Natural Gas Act, 52 Stat. 821, 15 U. S. C. § 717 (1958 ed.).1 26 F. P. C. 606, rehearing denied, id., at 840. The Court of

*369Appeals reversed, one judge dissenting. 323 F. 2d 190. The case is here on a writ of certiorari. 377 U. S. 951.

We said in Connecticut Co. v. Federal Power Comm’n, 324 U. S. 515, 529, “Federal jurisdiction was to follow the flow of electric energy, an engineering and scientific, rather than a legalistic or governmental, test.” And that is the test we have followed under both the Federal Power Act and the Natural Gas Act, except as Congress itself has substituted a so-called legal Standard for the technological one. Id., at 530-531. In Interstate Natural Gas Co. v. Federal Power Comm’n, 331 U. S. 682, 687, we considered the anatomy of the pipeline system to discover the channel of the constant flow; again in Federal Power Comm’n v. East Ohio Gas Co., 338 U. S. 464, 467; and most recently in Federal Power Comm’n v. Southern Cal. Edison Co., 376 U. S. 205, 209, n. 5. The result of our decisions is to make the sale of gas which crosses a state line at any stage of its movement from wellhead to ultimate consumption “in interstate commerce” within the meaning of the Act.

Attempts have been made by one convention or another to convert a local transaction into one of interstate commerce (Sprout v. South Bend, 277 U. S. 163; Superior Oil Co. v. Mississippi, 280 U. S. 390) or to make a segment of interstate commerce appear to be only intrastate (Baltimore & Ohio R. Co. v. Settle, 260 U. S. 166). But those attempts have failed. Similarly, we conclude that when it comes to the question what gas is for “resale” the present contracts should not be able to change the jurisdictional result.

The fact that a substantial part of the gas will be resold, in our view, invokes federal jurisdiction at the outset over the entire transaction. Were suppliers of gas and pipeline companies free to allocate by contract gas from a particular source to a particular use, havoc would be raised with the federal regulatory scheme, as it was con*370strued arid applied in Phillips Petroleum Co. v. Wisconsin, 347 U. S. 672. A pipeline would then be able to discriminate in favor of its “nonjurisdictional” customers. Moreover, a pipeline compariy by a contract clause could immunize a particular supplier from the reach of federal regulation2 as defined by Phillips Petroleum Co. v. Wisconsin, supra. There would be created in those and in other ways an “attractive gap” in the federal regulatory scheme (Federal Power Comm’n v. Transcontinental Gas Pipe Line Corp., 365 U. S. 1, 28) which the producing States might have little incentive to close, since the gap would often involve either lower costs to intrastate customers or else merely higher pipeline costs which ultimately would be reflected in rates paid by consumers in other States. Whether cases could be conjured up where in spite of original commingling there might be a separate so-called nonjurisdictional transaction3 of a precise amount of gas not-for-resale4 within the meaning of the Act is a question we need not reach.

*371Finally it is said that the Commission should draw the appropriate lines between “jurisdictional'’ and “nonjuris-dictional” sales through the use of its rule-making power. But we cannot say. that the ádjüdicatory process is not an appropriate method, for drawing the line case-by-case (United States v. Public Utilities Comm’n, 345 U. S. 295) as in a,''host of other administrative determinations. The Commission has acted responsibly in this situation and its decision must be upheld.

Reversed.

Mr. Justice White took no part in the consideration or decision of these cases.

dissenting.

Today’s decision furnishes a too-ready answer to an intricate, problem of administrative regulation. It re7 fleets the sort of decision that is to be expected when the Court is willing to make a bare choice between two unrefined points of view as to regulatory method, without first being informed by the regulating agency concerned as to its evaluation of the competing factors — something that is indispensable to achieving a well-balanced solution of a problem such as this. The respective positions of the parties here each possesses the capacity to frustrate the scope of natural gas regulation ordained by the Congress. The Commission’s molecular theory, accepted by the Court with undefined reservations, results in expanding the regulatory scheme by sweeping within the Commission’s authority gas that has not been supplied or used for interstate resale (“nonjurisdictional” gas). The respondents’ contract-allocation position, on the other hand, might serve to contract the legitimate scope of regulation by interfering with the ability of the Commission to deal with gas restricted under a supply contract to *372“not-for-sale,” but which has been actually used by the pipeline-purchaser for interstate resale (“jurisdictional” gas).

Whether or not there is a middle ground that would more closely fulfill the purposes of the Natural Gas Act than either of the proposals now before us is something that this Court, is not competent to assess without expert guidance from the Commission, and we have been given none. Lacking this, I am unwilling to accept at this juncture the position of either party to this litigation. I think the Court should decline to pass upon these cases until the Commission has first illumined the regulatory problems involved through an appropriate exercise of its rule-making powers.1

The complexity and elusiveness of the matters with which we are asked to deal are best exposed from the vantage point of this Court by considering some of the questions to which allocation contracts in varying contexts give rise.

The Commission has, at least until this case, accepted the proposition that a single supplier to a pipeline may allocate by contract between the amount of gas used for jurisdictional purposes and the amount used nonjurisdic-tionally. For example, in City of Hastings v. Federal Power Comm’n, 221 F. 2d 31, a pipeline company sold gas to the city through one pipeline under two contracts, one covering the gas to be resold by the city, and the. other gas to be used by the city in its own plants. Although the gas was mingled in the common pipeline, the allocation was approved, and the latter gas was, without more, considered not spbject to Commission regulá-tión. A similar situation was presented in United States v. Public Utilities Comm’n of California, 345 U. S. 295, where a power company sold electricity to the Navy for *373use in its power plants and also for resale to dependent families.^- The absence of any allocation was fatal in that case, but the Court recognized that a different question would be presented if there had been two separate transactions. 345 U. S., at 316-318.

The result does not change when two or more suppliers are involved, provided that the allocation of nonjurisdic-tional gas is prorated among all of the suppliers. For example, if a pipeline company consumed 30% of its total volume of gas in its own plants, and sold 10% of the total volume in the State of production, each supplier could allocate 40% of its gas supply to nonjurisdictional use. Such was essentially the case in North Dakota v. Federal Power Comm’n, 247 F. 2d 173, where the allocation was upheld with Commission approval. If these cases are accepted by the Court, two corollaries follow: since gas is a fungible commodity, the mingling of gas does not alone render ineffective for purposes of Commission jurisdic.tion the allocation contracts, although the molecular identification of the nonjurisdictional gas is destroyed; and the fact that the prices paid for nonjurisdictional gas 2 may affect the rate base for the jurisdictional gas, is also not a critical factor at this stage.3

*374The issue now before the Court arises only when some suppliers are allocating part or all of their gas to nonjuris-dictional use* but others are not. This issue could arise commonly in two contexts: if existing suppliers were allocating pro rata, and new suppliers were added which did not allocate, the addition of the new suppliers might be thought not to destroy the validity of the existing allocation contracts since the new suppliers might be satisfying an increase in the demand for jurisdictional gas.4 The converse situation is presented in this casé, where the new suppliers are attempting to allocate, and existing suppliers are not. One possible test in such cases might be to determine the source of the demand for the gas supplied to El Paso by Houston and Lo-Vaca. To modify the argument used by respondents, if a separate pipeline were constructed from the Coquat station (at which the gas enters the El Paso system) to the point along El Paso’s system where the outflow will increase, would the sale' be jurisdictional or not? If in fact El Paso has formerly been using the same amount of gas in its compressors that it intends to use in the future, then the purpose of the Lo-Vaca allocation will be merely to release for interstate sale — to satisfy, the interstate demand— gas from other suppliers which formerly was used for non-jurisdictional purposes.

The record before us does not answer the question put. There is some indication that El Paso intends to construct new compressor plants, and may have to use more non-jurisdictional gas at its existing plants to handle the added gas received from Lo-Vaca under the unrestricted contract. Such a use would satisfy a nonjurisdictional de*375mand. However, there is also evidence that in fact El Paso’s consumption for nonjurisdictional purposes will remain constant, and that Lo-Vaca’s supplies will be used to satisfy an increased demand from interstate consumers. The fact that Lo-Vaca gas purportedly replaces the compressor gas supplies formerly furnished by other suppliers, thus releasing that gas. for interstate resale, should not defeat Commission jurisdiction under this analysis.

Another possible standard which suggests itself would, be to determine the probable percentages of gas from each supplier which will be used for nonjurisdictional purposes, and only permit each supplier to allocate by contract to nonjurisdictional use his pro rata share of the. total estimated nonjurisdictional gas. For example, if we suppose a pipeline running from the Gulf coast of Texas through New Mexico into California, as does the El Paso system, then each supplier should determine what percentage of the total volume of gas. flowing west from the point of its input will be ultimately used for a¡ nonjurisdictional purpose: It would then be mathematically probable that his gas would be used for nonjurisdictional purposes in the same percentage, and he could allocate that amount by contract, subject to change should new supplies be added to the system.5

I recognize, of course, that there may be pitfalls in both of these possible methods,, and that there may be other, formulae that are preferable to either. I have ventured *376them only as support for my belief that the Commission’s molecular theory, which in the name of protecting the Commission’s jurisdiction in reality involves a judicial expansion of its authority, should not be accepted until the Commission, after due exploration in a rule-making proceeding, is able to satisfy this Court that no other feasible method — more particularly no modification of the respondents’ contract-allocation theory — exists that would better fit the boundaries of the Commission’s jurisdiction as fixed by Congress.

It is undoubtedly true that normally an administrative agency may decide for itself whether to proceed in a.given field of its regulatory functions through the promulgation of general rules6 or by the process of case-by-case adjudication.7 This Commission from the outset has usually proceeded, with the Court’s approval,8 in developing its procedures by the adjudicatory process. Nevertheless, there are good reasons why the rule-making power appears to be the more promising avenue of approach in this instance. First, the adjudicatory process has not yielded any satisfactory basic principle to serve as a point of departure for judicial assessment of cases of this kind, or indeed for a consistent administrative approach;9 even in this litigation the Commission’s position is far from clear as to what room, if any, there may be for restrictive allocation contracts. Second, the gas industry is entitled to know the fundamental ground rules by which it should *377conduct itself in this regard with some degree of predictability, as witness the situation of these respondents whose good faith in the transactions giving rise to this litigation has not been impugned in any way. Third, that unlike the line of cases in which agency jurisdiction is conceded,10 here the Commission should not be permitted to adopt a theory which expands its jurisdiction beyond statutory limits11 without full hearings and the formulation of a rule interpreting its jurisdiction in this area which conforms to the jurisdictional limits of § 1 (b) of the Natural Gas Act. Fourth, because these matters are fraught with technical “perplexities, both geological and economic,” Railroad Comm’n of Texas v. Rowan & Nichols Oil Co., 311 U. S. 570, 574, the informed expertise of the Commission is a necessary adjunct to satisfactory judicial resolution of particular cases. “Had the Commission, acting upon its experience and peculiar competence, promulgated a general rule of which its order here was a particular application, the problem for our consideration would be very different.” Securities & Exchange Comm’n v. Chenery Corp., 318 U. S. 80, 92. The courts have a right to the informed judgment of the Commission before acting further in this presently opaque area.

I would vacate the judgment of the Court of Appeals and remand the case to the Commission for further proceedings after the promulgation of interpretive rules to. cover this, and like cases.12