RADER, Circuit Judge.
The General Accounting Office (GAO) denied Advanced Data Concepts, Inc.’s (ADC’s) bid protest. ADC appealed the contract award to the United States Court of Federal Claims, which upheld the bid process and denied the protest. Because the trial court correctly found that the United States Department of Energy (DOE) did not abuse its discretion, this court affirms.
I.
DOE solicited proposals in April 1997 for the on-site delivery of specialized support services at its Office of Declassification in Germantown, Maryland. The solicitation notified the bidders that DOE would award the bid to the offer “most advantageous to the Government” based on evaluation of technical and cost criteria. The solicitation specified that the “technical proposal is of greater importance than the cost proposal.” The solicitation then set forth various categories of technical evaluation in order of importance. One important technical evaluation criterion was past performance. The solicitation asked each bidder to identify references which might supply evidence of past performance. In the event these references did not respond, the past performance rating for that reference would be neutral.
The solicitation also required DOE to conduct discussions with “all responsible offerors whose proposals [fall] ... within the competitive range.” Following the discussions, the offerors could submit a best and final offer (BAFO) before the contract award.
DOE received offers from ADC, DynCorp EENSP, Inc. (doing business as DynMeridian), and two other companies. DynMeridian had been performing the required services for DOE since 1994 on a contract which was to expire in 1999. DOE solicited new bids because funds allocated to the original contract had been exhausted.
DOE awarded the contract to DynMeridian without conducting the mandated discussions with the other bidders, provoking ADC to file a protest with the GAO. In response, DOE rescinded the award and reopened the procurement. The new procurement required DOE to determine a competitive range of offerors, to open discussions with bidders on written questions, and to evaluate BAFOs after the discussions.
Following this process, DOE again awarded the contract to DynMeridian, even though DynMeridian’s bid of $15.9 million was about 26% higher than ADC’s bid of $12.6 million. DOE justified the award on the basis of DynMeridian’s point evaluation score which, at 970/1000, was considerably better than ADC’s score of 645/1000. DOE also noted that none of ADC’s past-performance references responded to DOE’s request for information.
In a second bid protest to GAO, ADC raised ten claims for relief. While rejecting most of the claims, the GAO did find that DOE had made errors in applying its evaluation criteria and in evaluating ADC employees’ experience. With these errors corrected, ADC’s score rose to 665/1000. Because the correction improved ADC’s score only a little, GAO found no prejudice in DOE’s evaluation errors. Therefore, GAO upheld the DOE bid process.
ADC appealed to the United States Court of Federal Claims. Before the Court of Federal Claims, ADC claimed that: (1) DOE failed to rate plaintiffs proposal on the “availability of personnel” and “key personnel” subcriteria in accordance with the weighting in the solicitation; (2) DOE double-counted a weakness in plaintiffs proposal by rating both key and non-key personnel as “classification professionals,” (3) DOE failed to contemporaneously evaluate the effect of DynMeridian’s offer of uncompensated overtime; (4) DOE failed to contemporaneously assess the strengths and weaknesses of DynMeridian’s proposal; (5) DOE improperly conducted an ad hoc evaluation of plaintiffs past performance; (6) DOE failed to assess plaintiffs past performance under the statutory and regulatory framework for collecting such information; (7) DOE failed to afford plaintiff meaningful discussions; (8) DOE failed to notice that plaintiff had eliminated references to critical weapon design information; (9) DOE failed to contemporaneously document the technical evaluation, the source selection decision, and the procurement transaction; and (10) DOE breached its implied-in-fact contract to fairly evaluate plaintiffs proposal. See Advanced Data Concepts, Inc. v. United States, 43 Fed.Cl. 410, 413 (1999) (ADC I).
The Court of Federal Claims found that DOE had erred in its weighting of “availability of personnel” and “key personnel” in the submissions of all the bidders. The trial court determined that DOE weighted these criteria contrary to the terms of the solicitation. The Court of Federal Claims also found that DOE improperly weighted the document-classification experience of both “key” and “non-key” personnel. After correcting these errors in both DynMeridian’s and ADC’s proposals, ADC’s overall score rose to 711.6/1000. DynMeridian’s score did not change.
The Court of Federal Claims also gave ADC the benefit of the doubt concerning its claim that DOE had not properly evaluated its past performance. The trial court did not actually find for ADC on this point, but calculated that even if ADC had a perfect score on this element, its overall score would only have reached 761.6/1000. In other words, even after giving ADC the benefit of maximum scores in all of the quantifiable elements of the evaluation which it disputed, its technical evaluation score was 27% lower than DynMeridian’s. Its price was 26% lower.
Finding no merit in the remainder of ADC’s claims, the Court of Federal Claims concluded that the 27% rating difference was “particularly significant in light of the solicitation’s emphasis on technical factors over price ... and ... outweighs plaintiffs 26% cost advantage.” ADC I, 43 Fed.Cl. at 422. Therefore, the Court of Federal Claims detected no prejudice to ADC and granted summary judgment for DOE. See ADC I, 43 Fed.Cl. at 410.
II.
To prevail in its appeal to this court, ADC must show not only that significant errors occurred in the procurement process, but also that the errors were prejudicial. In other words, ADC must show that, if DOE had made no errors, “there was a reasonable likelihood that [it] ... would have been awarded the contract.” Data General Corp. v. Johnson, 78 F.3d 1556, 1562 (Fed.Cir.1996); see also Statistical Inc. v. Christopher, 102 F.3d 1577, 1581 (Fed.Cir.1996); CACI, Inc. v. United States, 719 F.2d 1567 (Fed.Cir.1983). Prejudice is a question of fact in a post-award bid protest over which the Court of Federal Claims has jurisdiction under 28 U.S.C. § 1491(b)(1) (1996).
In post-award bid protests, courts apply the standard of review for agency action under the Administrative Procedure Act, 5 U.S.C. § 706 (1994). See 28 U.S.C. § 1491(b)(4) (1996). Thus, § 706 directs a reviewing court to
hold unlawful and set aside agency action, findings, and conclusions found to be—
(A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law [or] ...
(E) unsupported by substantial evidence in a case subject to sections 556 and 557 ... or otherwise reviewed on the record of an agency hearing provided by statute----
5 U.S.C. § 706 (1994). See also Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 283, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974). The Court of Federal Claims evaluated directly the agency’s decision, not the GAO decision. See ADC I, 43 Fed.Cl. at 416; Burroughs Corp. v. United States, 223 Ct.Cl. 53, 617 F.2d 590, 597 (1980). The Court of Federal Claims confined its review to the administrative record, i.e., the record before the decision maker at the time of the final award. See ADC I, 43 Fed.Cl. at 416; Camp v. Pitts, 411 U.S. 138, 142, 93 S.Ct. 1241, 36 L.Ed.2d 106 (1973).
This court reviews without deference the Court of Federal Claims’ application of the summary judgment standard. See, e.g., Goodyear Tire & Rubber Co. v. Releasomers, Inc., 824 F.2d 953, 954 (Fed. Cir.1987). Therefore, this court also reviews the administrative record before DOE. On that administrative record, the parties present no factual disputes nor conflicting record evidence. Rather, the question before this court on appeal is the trial court’s finding of no prejudice in DOE’s evaluation errors.
The Court of Federal Claims correctly performed its review in this case under § 706(2)(A), the “arbitrary or capricious” standard. See ADC I, 43 Fed.Cl. at 416. This court reapplies that standard on review. The § 706(2)(A) “arbitrary and capricious” standard applies to bid protests under 28 U.S.C. § 1491(b)(4) reviewed in the absence of a hearing. Bid protests in the absence of a hearing do not present an agency record derived from a hearing provided by statute or under 5 U.S.C. § 556 or 5 U.S.C. § 557. Therefore the “substantial evidence” standard does not apply. See Camp, 411 U.S. at 141, 93 S.Ct. 1241. The arbitrary and capricious standard applicable here is highly deferential. This standard requires a reviewing court to sustain an agency action evincing rational reasoning and consideration of relevant factors. Bowman, 419 U.S. at 285, 95 S.Ct. 438.
As the Court of Federal Claims properly noted, DOE’s evaluations of the offers in the bid were reasonable and complied with the solicitation. Even adjusting the scores of ADC and DynMeridian to give ADC the benefit of every inference and potential factual dispute does not bring ADC’s technical evaluation scores anywhere near the proximity to DynMeridan’s that would be necessary to render the DOE evaluation arbitrary and capricious. After all, the solicitation gave all bidders notice that DOE would give more weight to technical proposals than to cost proposals. Accordingly the trial court correctly determined, under the applicable standard of review, that ADC suffered no prejudice from any DOE evaluation errors. Because this court detects no abuse of discretion in the DOE evaluation, it affirms the Court of Federal Claims’ grant of summary judgment to DOE.
COSTS
Each party shall bear its own costs.
AFFIRMED.