Opinion for the court. filed by Circuit Judge J. SKELLY WRIGHT.
J. SKELLY WRIGHT, Circuit Judge:
Petitioner Arizona Public Service Company, believing that rates charged by a number of railroads for delivery of oil to its generating facilities were unreasonable, petitioned the Interstate Commerce Commission to order the railroads to institute new, lower rates on the named routes. Under deregulatory legislation enacted by Congress in 1976 and 1980, the ICC can only investigate the reasonableness of a challenged rate if the railroads have “market dominance over the transportation to which the rate applies.” 49 U.S.C. § 10709(b) (Supp. V 1981.) (All references to Title 49 of the United States Code are to Supplement V 1981). The Administrative Law Judge in this case found requisite market dominance, see Arizona Public Service Co. v. Atchison, Topeka & Santa Fe R. Co., et al., February 15, 1983 (ICC Docket No. 38088S), Joint Appendix (JA) 373 (hereinafter cited as “AU Op.”), but the ICC’s Review Board reversed the AU’s findings, see Arizona Public Service Co. v. Atchison, Topeka & Santa Fe Railway Co., et al., July 19, 1983 (ICC Decision No. 38088S), JA 482 (hereinafter cited as “Review Board Op.”). Petitioner then brought the instant petition to this court pursuant to Section 10 of the Administrative Procedure Act, 5 U.S.C. § 706 (1982), seeking to have the Review Board’s decision overturned.
Petitioner claims that it made the strongest possible showing of market dominance and that the Review Board’s reversal of the AU was arbitrary and capricious and was not supported by substantial evidence in the record. We agree that the Review Board largely ignored the evidence fully supporting a finding that the railroads did indeed have market dominance over the transportation to which the rate applies. We therefore vacate the Review Board’s decision and remand for further proceedings consistent with this opinion.
I. Background
Petitioner is a major utility that provides electricity for about 70 percent of the population of the State of Arizona. The AU’s decision in this case supplies useful and uncontroverted details concerning petitioner’s generating capacity. Because electricity cannot be stored, petitioner “must at all times maintain more than sufficient gener*309ating capacity on line and in reserve to produce enough power at any given moment to match the highest reasonably predictable peak load.” AU Op. at 10, JA 381. A three-level generating system accomplishes this task. Coal-burning plants handle the base load demand for electricity. These plants require some distillate fuel oil to ignite the coal and to stabilize the temperatures at which it is burned. In a usual year the demand for distillate oil for such purposes is approximately 80,000 barrels. Id. When the demand for electricity approaches the capacity of the coal-burning plants, steam units capable of burning either residual fuel oil (which has markedly different characteristics from distillate oil) or natural gas come on line. Finally, special combustion turbines are used for periods of peak demand.
Five of petitioner’s 40 electrical generating units use oil as their only fuel, and 17 more are equipped to burn either oil or natural gas. As a percentage of total generating capacity, petitioner uses a relatively small amount of oil. In 1981, 89.5% of petitioner’s electricity was generated by coal, 9.9% by natural gas, and 0.6% by oil. These figures do not tell the entire story, however. Petitioner operates under three constraints when it purchases fuel. First, it naturally seeks to generate electricity using the least expensive fuels available. Coal has for some time been vastly preferable to natural gas and oil on this count. For this reason petitioner makes maximum use of coal to meet base load demands. Second, petitioner must consider state environmental regulations that limit the maximum sulphur dioxide emissions of petitioner’s plants. Natural gas is a relatively clean fuel and thus is preferable to oil from this perspective; in addition, natural gas has been cheaper than oil at least since 1979. Petitioner accordingly now makes maximum use of natural gas to meet peak load demands. Third, petitioner must consider the availability of the various fuels. From 1972 until May of 1981 the federal government curtailed the supplies of natural gas to petitioner. The extent of curtailment varied during this period, but the general result of the curtailment was that petitioner had to increase purchases of oil to make up for the shortage of natural gas.
The oil that petitioner does use comes from a variety of sources. In part, this is because petitioner purchases from the least expensive supplier at any given time. Petitioner must also purchase only varieties of distillate and residual oils that, when burned, will remain within pollution limits set by state environmental regulations. Refineries produce varying grades and varieties of oil depending upon the characteristics of the crude oil supply at any given time, the demands of the overall market for particular product mixes, and the technical problems involved in refining oil. Much refined oil available at any given time will not meet environmental standards when burned in petitioner’s plants. When petitioner needs oil, it needs oil immediately and, in light of the limited supply of acceptable grade oil, often has relatively little choice about where to purchase; it simply must purchase the appropriate grades of oil from whichever refinery has sufficient quantities available.
Petitioner transports oil via pipeline, rail, and truck to its oil-burning plants. Its preferred mode of transportation is a pipeline running between pumping stations near Los Angeles and Phoenix. Residual oil may not be shipped by pipeline. For distillate oil the pipeline charges 79.7 cents per barrel, as opposed to the rail rate of $2.72 per barrel for the same transportation. AU Op. at 8, JA 380. But because pipeline capacity is quite limited and because most of petitioner’s generating units are not located along the pipeline, petitioner uses rail to supply most of its distillate oil. A small amount of residual and distillate oil moves by truck. Truck rates, however, are considerably higher than rail rates. A sample of prices in the record indicates that truck rates for residual oil are approximately 30% to 50% higher than rail rates, while truck rates for distillate oil are approximately 20% to 60% higher than rail rates. Id.
*310On March 27, 1981 petitioner filed with ICC a complaint against several railroads that transport distillate and residual fuel oil from various suppliers to Petitioner’s generating plants. The timing of the complaint is largely explained by Section 229 of the Staggers Rail Act of 1980, Pub.L. No. 96-448, 94 Stat. 1895 {codified at 49 U.S.C. § 10101a et seq.), which required a shipper or receiver to challenge the reasonableness of existing rail rates within 180 days of the effective date of the Act. See 49 U.S.C. § 10701a note. Petitioner’s timely complaint challenged the reasonableness of rail rates charged for transportation of oil between 52 origin/destination pairs, the origins being oil suppliers in neighboring states and the destinations being petitioner’s power plants. Under the regulatory scheme ICC can exercise jurisdiction to declare rates unreasonable only if the railroad has “market dominance” over the routes to which the challenged rates apply. 49 U.S.C. § 10709(b). The AKT who initially adjudicated petitioner’s complaint found such market dominance and evaluated the reasonableness of the rates. The ICC Review Board reversed the ALJ on the question of market dominance. Essentially the Review Board found that effective direct and indirect competition constrained the challenged railroads’ ability to charge excessive rates, and the Board therefore concluded that petitioner had not shown market dominance. Petitioner challenges this decision of the Review Board.
II. Analysis
A. The Legal Framework
Until 1976 the ICC exercised general supervisory jurisdiction over rates for rail transportation. “A ‘just and reasonable’ standard controlled all rates, rail users could challenge the reasonableness of rates, at any time, and the Commission had authority to determine what constituted a just and reasonable rate in each case.” Ford Motor Co. v. ICC, 714 F.2d 1157, 1158 (D.C.Cir.1983). Congress altered the statutory scheme to limit significantly ICC’s jurisdiction over rail rates in 1976 in the Railroad Revitalization and Regulatory Reform Act, Pub.L. No. 94-210, 90 Stat. 31 (4-R Act). In the 4-R Act Congress provided that ICC must find that the railroad has “market dominance over the transportation to which the rate applies” as a prerequisite to exercising its power to determine the reasonableness of the rate. 49 U.S.C. § 10709(b). Underlying this provision is the theory that if the railroad does not have market dominance competitive pressures will keep rail rates at a reasonable level. As Congress stated in the statute, “ ‘market dominance’ means an absence of effective competition from other carriers or modes of transportation for the transportation to which a rate applies.” 49 U.S.C. § 10709(a). ICC’s intervention is therefore appropriate only in situations where such competitive pressures are absent.
Congress again amended the statutory scheme with passage of the Staggers Rail Act of 1980, Pub.L. No. 96-448, 94 Stat. 1895 {codified at 49 U.S.C. § 10101a(l) et seq.). In Section 229 of the Staggers Act Congress decided that, even where the railroad had market dominance, as defined in the 4-R Act, a complainant could challenge rates existing on the effective date of the Staggers Act (October 1, 1980) only (with some exceptions not relevant here) for a period of 180 days after that date. See 49 U.S.C. § 10701a note; Ford Motor Co. v. ICC, supra, 714 F.2d at 1159; Baltimore Gas & Electric Co. v. ICC, 672 F.2d 146, 147-148 (D.C.Cir.1982). Congress enacted this provision because calculations called for in other sections of the Staggers Act required the existence of some base rate that could be accepted as legal and would itself not be subject to challenge. See 49 U.S.C. § 10707a (requiring a base rate to use in calculating the “zone of rail carrier rate flexibility”); Baltimore Gas & Electric Co. v. ICC, supra, 672 F.2d at 148; New England Power Co. v. United States, 693 F.2d 239, 242 n. 5 (1st Cir.1982). In this case petitioner brought a timely challenge to certain railroad rates in existence as of the effective date of the Staggers *311Act. Before evaluating the reasonableness of the challenged rates, ICC thus had to make a threshold determination whether the railroads had market dominance over the transportation to which the rates applied.1
To implement the 4-R Act in 1976 ICC promulgated a rule stating that two forms of competition would be considered in deciding such market dominance questions: (1) “intramodal” competition, or competition from other rail carriers in the same transportation route serving the same origin/destination pair as that for the disputed rate and (2) “intermodal” competition, or competition from other modes of transportation for transporting the commodity between the same origin/destination pair. See Ex Parte No. 320, Special Procedures for Making Findings of Market Dominance, 353 ICC 874 (1976). If effective intermodal or intramodal competition were present, a carrier would be held not to have the requisite market dominance and the Commission would not rule on the reasonableness of its rates. In Atchison, Topeka & Santa Fe R. Co. v. ICC, 580 F.2d 623 (D.C.Cir.1978), this court upheld the Commission’s 1976 rules for determining market dominance. After the passage of the Staggers Act, ICC returned to the issue of how to determine market dominance in Ex Parte No. 320 (Sub-No. 2), Market Dominance Determinations and Consideration of Product Competition, 365 ICC 118 (1981) (hereinafter cited as Market Dominance Determinations). For present purposes, the most important result of the Commission’s reconsideration was the conclusion that two additional forms of competition would be considered in determining whether a railroad had market dominance on a given route: (1) “geographic” competition, or competition between the challenged railroad and other carriers delivering the same product to the same destination from alternative origins, and (2) “product” competition, or competition between the challenged railroad and other carriers delivering substitute products to the same destination, irrespective of origin. See id. at 129, 132,133. The Commission’s new rules included detailed guidelines specifying the kinds of evidence that protestants should submit on each of these points. Thus under current practice, if the Commission’s examination of these four forms of competition — intramodal, intermodal, geographic, and product — indicates that the challenged railroad has effective competition on the challenged route, the Commission finds that the carrier does not have market dominance and the reasonableness of the challenged rates is not subject to ICC scrutiny.
The new rules first received judicial review on December 8, 1982 in Western Coal Traffic League v. United States, 694 F.2d 378 (5th Cir.1982), rev’d, 719 F.2d 772 (1983) (en banc), cert. denied, — U.S. -, 104 S.Ct. 2160, 80 L.Ed.2d 545 (1984), in which a Fifth Circuit panel ruled that ICC could not, consistent with the 4-R Act, take product and geographic competition into account in determining whether a railroad had market dominance. When the ALJ decided that the railroads in this case did have market dominance, he was proceeding in accord with the Fifth Circuit panel decision and accordingly took only intramodal and intermodal competition into account. After his opinion was issued on February 15, 1983, the en banc Fifth Circuit vacated the panel decision and set the case down for rehearing by the full court. On November 14, 1983 the full court reversed the panel decision and affirmed ICC’s decision to consider evidence of geographic and product competition. Western Coal Traffic League v. United States, 719 F.2d 772 (5th Cir.1983) (en banc). Thus the Review Board in its decision of July 19, 1983 — the decision before us today — considered all four types of competition in reaching the conclusion that the railroads involved did not have market dominance.
This court has not yet decided whether the new ICC rules comport with the intent *312of Congress when it adopted the “market dominance” standard for the ICC. In this case, however, all parties have assumed in their submissions to us that the new rules are valid. We will therefore assume without deciding that all four types of competition may be considered in reaching a market dominance determination. Thus in this case we review ICC’s resolution of whether, taking all four forms of competition into consideration, there is “effective competition from other carriers or modes of transportation for the transportation to which a rate applies.” 49 U.S.C. § 10709(a).
B. Standard of Review
Section 10 of the Administrative Procedure Act, 5 U.S.C. § 706 (1982), governs our review of ICC’s decision that the railroads do not possess market dominance over the transportation to which the challenged rates apply. We must vacate the decision if it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law * * * [or] unsupported by substantial evidence.” Id. § 706(2)(A) & (E). Though our review is “searching and careful,” the ultimate standard of review is “narrow.” “The court is not empowered to substitute its judgment for that of the agency.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971); see Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285-286, 95 S.Ct. 438, 441-442, 42 L.Ed.2d 447 (1974). Our review ensures that the agency has engaged in the reasoned decisionmaking essential to informed and evenhanded implementation of public policy, Cross-Sound Ferry Services, Inc. v. ICC, 738 F.2d 481, 483 (D.C.Cir.1984), and that the agency’s findings and conclusions are supported by substantial evidence in the record as a whole. Universal Camera Corp. v. NLRB, 340 U.S. 474 (1951).2
C. Evaluation of the Review Board’s Market Dominance Determination
Cognizant of these principles of review, we evaluate the Review Board’s decision in this case that petitioner had not shown market dominance. We will evaluate the Review Board’s conclusions with respect to each of the four types of competition considered under Market Dominance Determinations, supra, in order to determine whether these conclusions are the product of reasoned decisionmaking and are supported by substantial evidence in the record as a whole.
1. Intramodal competition. The Review Board noted that “[the railroads] do not seriously contest market dominance on the ground of intramodal competition.” Review Board Op. at 5, JA 486. Transportation of oil to petitioner is “handled by the various defendants based on geographic considerations,” id., in that transportation between a given origin and a given destination is handled by the railroad that serves *313that route. Therefore, the Review Board appropriately refused to find any intramodal competition that would be sufficient to keep the railroads from having market dominance on the-several routes being challenged.
2. Intermodal competition.
(a) Truck transportation. The ALT found that trucks did not provide effective intermodal competition with railroads for the shipments of oil at issue in this case. The Review Board reversed the ALT’S finding for the following reasons:
In [a prior case], it was concluded that although the railroads transported more than 90 percent of the subject traffic, motor carrier competition was, nevertheless, feasible. That is similar to the instant situation, where, as pointed out above, for the limited fuel oil movement that is utilized by the complainant, the [railroads] handle 88 percent and motor carriers transport 12 percent. Therefore, as in [the prior case], we conclude that motor carrier competition is feasible. * * * For competitive pressures to be present, a competing mode would not need to be capable of handling substantially all or even a majority of the subject traffic. * * 11 We conclude that complainant has failed to show that motor carriers do not exert competitive pressure upon rail carriers for the traffic in issue.
Review Board Op. at 5, JA 486 (citation and footnote omitted).3
The Review Board’s “reasoning” here seems to go as follows: Petitioner showed only that railroads carry a vastly greater percentage of the oil traffic than do trucks. Petitioner’s use of railroads to transport its oil is consistent with the hypothesis that trucks provide effective competition for the railroads. This can be the case if, for example, truck rates are just slightly above rail rates and railroads can therefore maintain their market share only if they do not raise rates and provide services equal to or better than trucks. Therefore, petitioner has failed to prove that the railroads have market dominance in transporting oil on the challenged routes.
This argument is insufficient to support the Review Board’s conclusion. ICC’s own guidelines state that “[effective competition from motor carriage may be deduced from,” inter alia, “the amount of the product in question that is transported by motor carrier.” Market Dominance Determinations, supra, 365 ICC at 133. Petitioner submitted probative evidence of precisely the kind that the Commission’s rules encourage complainants to submit, and the Commission may not refute that evidence with the bald statement that such evidence is not absolutely conclusive on the issue. The reasoned decisionmaking requirements of Section 10 of the APA preclude setting this kind of trap for an unwary litigant.
Moreover, the Review Board’s opinion inexplicably ignores considerable — and highly significant — evidence concerning market power that petitioner introduced in addition to the evidence as to relative market share of railroads and trucks. Most important was the evidence summarized above that truck rates were 30% to 50% higher than rail rates for transporting residual oil over the same routes, and 20% to 60% higher for distillate oil. The Commission’s guidelines state that evidence of effective competition may include “the transportation costs of the rail and motor carrier alternatives.” Id. Such evidence is highly probative on the issue whether trucks provide effective competition for railroads for the given routes. At the core of the “effective competition” standard is the idea that there are competitive, market pressures on the railroads de*314terring them from charging monopoly prices for transporting goods. Of course, any such effective competition will always he relative to a particular price that the railroads charge. At some point the availability of an alternative such as the horse and buggy or even people carrying oil in buckets theoretically prevents railroads from raising their rates beyond an outer bound. But the mere existence of some alternative does not in itself constrain the railroads from charging rates far in excess of the just and reasonable rates that Congress thought the existence of competitive pressures would ensure. All record evidence in this case shows that truck rates are much higher than railroad rates for comparable services, and there is no suggestion in this record that the truck rates are higher because of any superiority in truck transportation of oil. Indeed, the record shows truck transport to be inferior due to the limited truck off-loading facilities at petitioner’s plants. AU Op. at 14, JA 386. Nothing in the record tends to suggest that in the special circumstances of the particular market the truck rates exert significant competitive pressure on railroad rates, and the Review Board does not draw our attention to any such special circumstances. Thus, even given the requisite degree of deference under the substantial evidence test, this record will not support the conclusion that trucks provide effective competition for the railroads at current price levels.
(b) Pipeline transportation. Intervenor railroads urge that, even if the Review Board were incorrect concerning intermodal competition from trucks, the challenged railroads nonetheless face effective inter-modal competition over the challenged routes from pipelines. In the section of its decision labeled “Background” the Review Board did briefly mention the fact that “fuel oil by pipeline (mainly distillate oil) is much cheaper than comparable quantities by rail.” Review Board Op. at 3, JA 484. See also AU Op. at 8, JA 380 (pipeline charge is 79.7 cents and rail rate is $2.72 to transport a barrel of oil over the identical route). It is far from clear that the Review Board relied on the existence of the pipeline in reaching its determinations; it is not mentioned in the Board’s discussion of intermodal competition in the section of the opinion labeled “Discussion and Conclusions.” Thus intervenors’ argument must be viewed as an impermissible post hoc rationalization of ICC’s decision.
Even if ICC had relied on such a line of reasoning, however, we find no evidence in the record to support the conclusion that the pipeline provides competition for the challenged transportation. First, all parties agree that, contrary to the Review Board’s statement, residual fuel cannot be transported via the available pipeline; the pipeline could therefore only provide inter-modal competition — if at all — for transportation of distillate oil. Second, the AU’s careful opinion notes several reasons why pipeline transportation is not generally available to petitioner: “No pipeline service extends to any of the destinations of the assailed rates except the Phoenix power plant,” AU Op. at 8, JA 380, and “[petitioner] has shipped to Phoenix as much of its distillate requirements via the pipeline as the capacity allotments by [the pipeline company] to [petitioner] would permit.” Id. Third, the difference in rates itself suggests that the challenged rail transportation is relatively unaffected by pipeline competition, for petitioner would presumably not use the railroads at all if the pipeline were available to supply service at a vastly lower price.
In short, petitioner introduced convincing evidence, of precisely the kind that the Commission’s own guidelines call for, to the effect that — under current market conditions — motor carriers provided no effective competition whatever for the challenged transportation. In rejecting this evidence the Review Board pointed to nothing in the record that might reasonably be viewed as factual support for the position that intermodal competition constrains prices in this case. Thus the Review Board’s conclusion with respect to intermodal competition must be viewed as arbi*315trary and capricious and lacking substantial supporting evidence in the record.
3. Product competition. The Review Board held that “[a] more complete and total picture of product competition and, in fact, substitution, is difficult to imagine. Beyond question, product competition exists for this market.” Review Board Op. at 6, JA 487. This conclusion is derived from the following discussion:
Product competition exists when a receiver or shipper can use a substitute for the product covered by the rail rate in issue. * * * The record here shows beyond doubt that product competition is present and controlling. Of the remaining 23 power plants still involved in the complaint, 4 of them use coal (except oil is used to a minor extent for igniting the generators) and 17 use and have used natural gas exclusively, for periods represented on this record as varying from 3 to 6 years. Only 2 of the power plants depend on oil as their fuel.
Id. (citation omitted). The Commission now concedes, as it must, that the Review Board made a factual mistake in finding that only two plants depend on oil as their only fuel. In fact, petitioner operates five plants that depend solely on oil. See brief for respondents at 19 n. 20. The question remains, nonetheless, whether the Review Board’s holding that product competition exists “beyond doubt” was a product of reasoned decisionmaking based on the current record.
According to ICC’s definition, “[p]roduct competition occurs when a receiver or shipper can use a substitute(s) for the product covered by the rail rate.” Market Dominance Determinations, supra, 365 ICC at 128. The fact that petitioner operates plants that burn coal and natural gas does not lead ineluctably to the conclusion that coal and natural gas provide product competition for oil. Such competition would only exist if, among other things, petitioner could readily substitute coal or natural gas for oil and if these alternative fuels were available in sufficient supply to restrain oil price increases (which of course include the railroads’ cost of transporting oil). Petitioner introduced evidence — largely credited by the AU — on both substitutability and availability. The Review Board shows no indication of having looked at the record (or even the AU’s decision) in this case closely enough to have been aware of this evidence.
(a) Coal as a substitute. We need not expend a great deal of time on the possibility that coal provides competition for oil in the context of this case, for we are unable to find any evidence that coal is in any way an acceptable substitute. To be sure, it is the fuel that petitioner uses to meet its base load requirements, see AU Op. at 9, JA 381, and we have no doubt that there may be many contexts in the national economy in which coal and oil are equally acceptable as fuel sources. However, the Review Board was adjudicating this case based on evidence submitted by these parties. In the context of this case oil is used for two purposes: (1) to stabilize combustion in the coal plants and (2) to meet peak load demands. See AU Op. at 9-10, JA 381-382. There was no evidence that coal itself can be used to stabilize combustion in the coal-fired plants. And .there was no evidence that petitioner, having made the capital investment necessary to use oil or natural gas for the peak load plants, could use coal as a substitute to meet peak load demands. Therefore, petitioner demonstrated that its demand could not be satisfied by coal, and we find no support for the Review Board’s evident conclusion to the contrary.
(b) Natural gas as a substitute. The question remains whether natural gas provided product competition for petitioner’s use of oil. Petitioner introduced significant evidence that natural gas did not provide product competition for oil.
(i) Substitutability to stabilize combustion in coal-fired plants. First, we are unable to find any evidence that petitioner could have used natural gas in place of the oil used to stabilize combustion in its coal plants. The Review Board’s only comment on this is that petitioner has “minimal fuel *316oil requirements to merely ignite the generators.” Yet the grounds for concluding that the 80,000 barrels per year used for this purpose, see AU Op. at 9, JA 381, are “minimal” and therefore in some way outside the protection of the Staggers Act are not clear in the Review Board’s opinion. Even if applicable laws would give ICC some discretion to decline to regulate rates on routes over which relatively small quantities of a commodity are shipped, and we have found nothing in the statute or legislative history to so indicate, the Commission would have to exercise this discretion in a reasoned fashion. The Review Board here seems to have entirely disregarded petitioner’s use of oil in its coal plants on the basis of pure administrative fiat.
(ii) Substitutability in peak load plants. Second, the AU found that petitioner could not substitute natural gas for oil in its peak load plants. The AU made it clear that petitioner’s demand for oil had some unique characteristics. The nature of the electrical generating industry is such that a utility must have capacity to generate electricity as it is demanded by its customers; a utility does not generally have the option of limiting its output to what it is able to produce at its “base load” plants. The utility must thus be able to respond to higher demand levels quickly by making use of its peak load plants. Whatever fuel is used in these plants must be readily available in sufficiently large quantities to meet peak demands. In addition, these fuels must be available in the specific types that will enable petitioner to comply with state environmental regulations.
Petitioner uses natural gas and oil to meet its peak load demands. Under current circumstances petitioner in fact relies on natural gas for the bulk of its demand because natural gas is readily available, is much less expensive, and comports more easily with environmental regulations. Under these circumstances, natural gas is so vastly preferable to oil as a fuel to meet peak load demands that petitioner burns oil only when it must — that is, when natural gas is for some reason unavailable to meet petitioner’s needs. In these situations natural gas cannot be considered a substitute for oil and the price of natural gas thus will not constrain the railroads’ ability to charge excessive rates for shipment of oil.4
Although the Review Board may be correct in its belief that natural gas is substitutable for oil in some places in our national economy, its conclusion that natural gas is substitutable in this case is pure conjecture, and cannot be squared with the facts found by the AU and left unrefuted by the Review Board. At the very least, evidence of product competition is “inconclusive.” Market Dominance Determinations, supra, 365 ICC at 130. Therefore, under the Commission’s own guidelines this factor is to receive “minimal consideration.” Id. Congress’ determination that economic concepts such as “market dominance” should be prerequisites to ICC regulation of a given rail rate was not a license for the Commission to adjudicate a party’s rights on the basis of “facts” originating in its own imagination. We therefore find the Review Board’s conclusion with respect to product competition to be arbitrary and unsupported by substantial evidence in the record.
4. Geographic competition. The Review Board found that the railroads faced geographic competition on the challenged routes. The Board noted that “the origins listed in the instant complaint are complainant’s principal sources but not its only sources. On account of the uncertainties surrounding oil production, complainant must maintain a broad spectrum of sources. Altogether, there are hundreds of refineries from which complainant can buy fuel oil.” Review Board Op. at 6, JA 487 *317(emphasis in original). Although the Review Board’s reasoning is far from clear, presumably it believed that these “hundreds” of other sources of oil in the country would be adequate for petitioner’s needs. It would follow that the availability of oil from these other sources would put competitive pressures on the railroads to hold down their own rates on the challenged routes.
We reject the Review Board’s conclusion. To be sure, there are indications in the record that petitioner has used oil from refineries not on the challenged routes in the past. If considered in isolation, that fact might provide some marginal support for the proposition that there is geographic competition in this case. However, our charge is to review the Commission’s action to determine if it is supported by substantial evidence on the record as a whole. See NLRB v. Universal Camera Corp., supra. We find that it is not.
Petitioner submitted evidence indicating that oil from other sources, though occasionally useful in the past, was insufficient to meet its needs, which include sporadic, urgent demands for oil of the particular grades and qualities that would comport with environmental regulations. The ALJ found that “[t]he origins involved in this proceeding are the principal sources available to [petitioner] of the residual and distillate fuel oils it needs and the rates assailed apply to the principal oil movements upon which it would have to depend when major burns or the greatest need for such oil occurs.” AU Op. at 5, JA 377.
The Review Board did not disagree with the ALJ’s findings. The Board instead attempted to capitalize on the AU’s use of the word “principal” to construct an entirely hypothetical argument that geographic competition nonetheless exists. The Review Board believed that “effective geographic competition is not contingent upon receivers being able to acquire identical tonnages from other sources.” Review Board Op. at 6, JA 487. In some circumstances, the Review Board claimed, availability of a relatively small amount of oil from other sources may cause a sufficient long-term loss to the railroad that it would be deterred from raising prices above a competitive level. The potential loss to the railroad over the long term may thus allegedly exceed any short-term gains.
We need not decide whether, as a matter of abstract economics, the Review Board was correct that the availability of a relatively small amount of oil from other sources could bring market discipline to the railroads. For even if this proposition were true as a matter of abstract economics, it could not be presumed true in every case in which some amount of oil, however small, were available from alternative sources. The extent of competitive pressure that these alternative sources would bring to bear would depend (at least) on (1) whether the alternative sources were served by the railroads involved in this case or by other carriers, (2) the extent of the monopolist’s profit that the railroads could reap by raising their prices, and (3) the amount of traffic that the railroads would lose by raising prices. The Review Board committed the error of accepting petitioner’s evidence (which was, by ICC’s own guidelines, clearly probative on the issue) at the same time as it rejected petitioner’s conclusions on the basis of the pure speculation that effective competitive pressures could exist. The APA requires that the Review Board’s conclusion instead be supported by substantial evidence, and, because we are unable to find such evidence in the record, we reject the Review Board’s conclusion that there was effective geographic competition in this case.
III. Conclusion
We hold that the ICC Review Board’s conclusion that the railroads did not have market dominance over the routes to which the rates petitioner challenges apply was not the product of the reasoned decision-making required by Section 10 of the APA. In the face of significant evidence of market dominance, conforming to the guidelines in ICC’s Market Dominance Determinations, supra, the Review Board can*318not justify its findings of a lack of market dominance on the basis of abstract speculation that has no foundation in the record of the case before it. Though recent congressional legislation has certainly shifted the Commission’s energies away from regulation of rail rates, Section 229 of the Staggers Act in particular embodies a compromise that authorizes petitioner’s challenge to the rates for oil shipped to its plants. The Review Board’s dismissal of this challenge “bear[s] the earmarks of a hasty route, not the measured passage from the old regime to the new for which Congress provided.” Ford Motor Co. v. ICC, supra, 714 F.2d at 1158. Accordingly, we vacate the Review Board’s decision and remand for further proceedings consistent with this opinion and with Section 10 of the APA.
Vacated and remanded.
. ICC also had to make a threshold determination that the challenged rates exceed a specified revenue-to-variable cost ratio. 49 U.S.C. § 10709. The Review Board agreed with the AU that the rates did exceed that ratio. Review Board Op. at 4, JA 485.
. Section 229(d) of the Staggers Act states that ”[t]he burden of proof in any proceeding under this section shall be on the complainant.” See 49 U.S.C. § 10701a note; see also Market Dominance Determinations, 365 ICC 118, 132 (1981) (protestant has burden of proving lack of effective competition). This case does not require us to determine the precise contours of the statutory burden that the complainant must bear. Clearly, however, Congress’ allocation of the burden of proof in this proceeding does not eviscerate the APA’s requirement that the agency decision be supported by substantial evidence in the record as a whole. See 5 U.S.C. § 706(2)(E) (1982). The Commission continues to have an affirmative obligation to present sufficient evidence to provide a "basis for informed judgment” about the relevant market. Brae Corp. v. United States, 740 F.2d 1023, 1043 (D.C.Cir.1984). Accordingly, the Commission must do more than merely speculate that there may be circumstances in which all of the protestant’s evidence may be true but in which the railroad still does not have market dominance. In Brae ICC had before it substantial evidence that railroads did not enjoy an excessive market share or abnormally high returns on variable costs in the generally relevant market, and that evidence made plausible the Commission's theoretical arguments that the railroads did not enjoy market power. Here, on the contrary, there is no substantial evidence supporting ICC’s finding of no market dominance. Whatever the allocation of the burden of proof, mere conjecture and abstract theorizing offered in a vacuum are inadequate to satisfy us that the agency has engaged in reasoned decisionmaking.
. The Review Board also found significant the fact that "the complainant has reduced the size of its fleet of tank cars, which it leases to the railroads for transportation of fuel oil, from 250 cars in 1979 to 186 cars in 1982, revealing a de-emphasis of rail transportation.” Review Board Op. at 5, JA 486. Absent any evidence of an increase in the amount of oil shipped by truck, however, this decline in the number of rail cars must be taken to indicate not a de-emphasis of rail transportation vis-a-vis truck transportation, but a decline in the overall use of oil. See Part I supra (petitioner has reduced use of oil in the past several years).
. The AU also found that natural gas supplies may be curtailed in the future, as supplies have been curtailed in the past, and that natural gas prices may increase significantly. AU Op. at 6-7, JA 378-379. Such developments, of course, would tend to increase petitioner’s existing dependence on oil. The Review Board does not dispute the AU’s findings of possible curtailment of natural gas supplies, of possible sharp increases in gas prices, or of petitioner’s need to use oil if either of these two eventualities comes to pass.