Gallant Ocean (Thailand) Co., Ltd. (“Gallant”) appeals from a final judgment of the United States Court of International Trade concerning its importation of frozen warmwater shrimp. The Court of International Trade affirmed a decision of the United States Department of Commerce (“Commerce”) to apply an adverse facts available (“AFA”) rate of 57.64% against Gallant. Gallant Ocean (Thailand) Co., Ltd. v. United States, 602 F.Supp.2d 1337 (Ct. Int’l Trade 2009). Because substantial evidence does not support the 57.64% AFA rate, this court vacates and remands.

I.

Commerce issues antidumping duty orders for imported merchandise that is sold in the United States below its fair value and materially injures or threatens to injure a domestic industry. See 19 U.S.C. § 1673 (2006). An antidumping duty reflects the amount by which the normal value exceeds the export price of a foreign exporter’s merchandise. 19 U.S.C. § 1673e(a)(l); 19 U.S.C. § 1677(35). This excess amount is also known as the “dumping margin.” The normal value is the price of the merchandise when sold for consumption in the exporting country. 19 U.S.C. § 1677b(l). If the imported merchandise is not sold in the exporting country, the normal value is the price at which the merchandise is sold for consumption in another similar exporting country or the United States. Id.

Commerce periodically reviews and reassesses antidumping duties. 19 U.S.C. § 1675(a). During its administrative review, Commerce requests information from the interested parties, including the foreign exporters of the subject merchandise. Upon a finding that an interested party refuses to cooperate with these information requests, Commerce “may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available.” 19 U.S.C. § 1677e(b). Therefore, Commerce can apply AFA rates against uncooperative parties. In calculating AFA rates, Commerce may rely on information derived from (1) the petition; (2) a final determination in the investigation; (3) any previous review; or (4) any other information in the record. Id. “When [Commerce] relies on secondary information rather than on information obtained in the course of an investigation or review, [Commerce] shall, to the extent practicable, corroborate that information from independent sources that are reasonably at [its] disposal.” 19 U.S.C. § 1677e(c).

II.

Gallant is a Thai exporter of shrimp. In December 2003, the Ad Hoc Shrimp Trade[*1322] Action Committee, a domestic committee, filed petitions with Commerce and the International Trade Commission (“ITC”), alleging that Thai and other foreign exporters were dumping frozen warmwater shrimp in the United States. Notice of Initiation of Antidumping Duty Investigations, 69 Fed.Reg. 3876 (Jan. 27, 2004). Based on the petition, Commerce calculated the Thai exporters’ dumping margin at 57.64%, as adjusted at the initiation of the less-than-fair value investigation. Id. at 3881. Thus, Commerce initially assigned an adjusted petition rate of 57.64% against Thai exporters.

After investigating the alleged dumping, Commerce issued an antidumping duty order on certain frozen warmwater shrimp from Thailand. Notice of Final Determination of Sales at Less than Fair Value and Negative Final Determination of Critical Circumstances: Certain Frozen and Canned Warmwater Shrimp from Thailand, 69 Fed.Reg. 76918, 76920 (Dec. 23, 2004). Commerce imposed dumping margins ranging from 5.91% to 6.82% against Thai exporters of shrimp. Notice of Amended Final Determinations of Sales at Less than Fair Value and Anti-dumping Duty Order: Certain Frozen Warmwater Shrimp from Thailand, 70 Fed.Reg. 5145, 5146 (Feb. 1, 2005) (“Final Antidumping Order”). Gallant did not participate in this initial investigation.

In April 2006, Commerce began its first administrative review of the antidumping duty order for the period covering August 4, 2004 through January 31, 2006. Notice of Initiation of Administrative Reviews of the Antidumping Duty Orders on Certain Frozen Warmwater Shrimp from Brazil, Ecuador, India and Thailand, 71 Fed. Reg. 17819 (Apr. 7, 2006). Commerce requested 145 Thai companies, including Gallant, to submit a quantity and value questionnaire. Id. at 17829. Gallant did not respond. Certain Frozen Warmwater Shrimp from Thailand: Preliminary Results and Partial Rescission of Anti-Dumping Duty Administrative Review, 72 Fed.Reg. 10669, 10673 (Mar. 9, 2007) (“Preliminary Results ”). In May 2006, Commerce asked Gallant to respond for the second time. Id. at 10673. Gallant again did not respond. Id.

In March 2007, Commerce found that Gallant had not acted to the best of its ability to cooperate with Commerce’s information requests and preliminarily assigned it a 57.64% AFA rate. Id. Commerce based the AFA rate on the adjusted petition rate. Id. at 10669-70. Commerce explained that it corroborated the adjusted petition rate with the transaction-specific margins calculated for the three mandatory respondents: Good Luck Product Co., Ltd. (“Good Luck Product”); Thai I-Mei Frozen Foods Co., Ltd. (“Thai I-Mei”); and Pakfood Public Co. Ltd. and its affiliated subsidies (collectively, “Pakfood”). Id. Specifically, both Good Luck Product and Pakfood had multiple transactions with dumping margins above the adjusted petition rate. Id. at 10673.

In September 2007, Commerce published the final results, continuing to assign a 57.64% AFA rate to Gallant and other uncooperative parties. Id. at 52069. Commerce assigned much lower dumping margins to cooperative parties: 10.75% for Good Luck Product; 2.58% for Thai I-Mei; 4.29% for Pakfood; and 4.31% for all other cooperative parties. Id.

Gallant challenged the 57.64% AFA rate as having no rational relationship to its commercial practices. Gallant Ocean, 602 F.Supp.2d at 1345. The Court of International Trade affirmed Commerce’s adverse[*1323] inference determination. Id. at 1346-52. Gallant now appeals to this court. This court has jurisdiction under 28 U.S.C. § 1295(a)(5).

III.

This court reviews a decision of the Court of International Trade concerning Commerce’s antidumping determination by reapplying the same standard of review used by the Court of International Trade. Tung Mung Dev. Co. v. United States, 354 F.3d 1371, 1378 (Fed.Cir.2004). This court upholds Commerce’s determinations unless they are “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” Alloy Piping Prods., Inc. v. Kanzen Tetsu Sdn. Bhd., 334 F.3d 1284, 1289 (Fed.Cir.2003) (internal citation omitted). Substantial evidence is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Micron Tech., Inc. v. United States, 117 F.3d 1386, 1393 (Fed.Cir.1997) (internal citation omitted). This court reviews the record as a whole, including any evidence that “fairly detracts from the substantiality of the evidence,” in determining whether substantial evidence exists. Id.

TV.

Commerce has broad discretion in making antidumping determinations. F. lli Di Ceceo De Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000). Commerce’s discretion is particularly great in the ease of uncooperative respondents. For example, Commerce can select from a list of secondary sources as a basis for its adverse inferences against uncooperative parties. See 19 U.S.C. § 1677e(b). “Commerce’s discretion in these matters, however, is not unbounded.” De Cecco, 216 F.3d at 1032.

An AFA rate must be “a reasonably accurate estimate of the respondent’s actual rate, albeit with some built-in increase intended as a deterrent to noncompliance.” Id. (emphasis added). The purpose of the AFA rate “is to provide respondents with an incentive to cooperate, not to impose punitive, aberrational, or uncorroborated margins.” Id. Therefore, although a higher AFA rate creates a stronger deterrent, Commerce may not select unreasonably high rates having no relationship to the respondent’s actual dumping margin. Id. Congress tempered the deterrent purpose with the corroboration requirement so as “to prevent the petition rate (or other adverse inference rate), when unreasonable, from prevailing and to block any temptation by Commerce to overreach reality in seeking to maximize deterrence.” Id.

In this case, Commerce incorrectly presumed that the adjusted petition rate was reliable in the face of much more reliable information and thus imposed an unreasonably high 57.64% AFA rate against Gallant. Cf. id. at 1033 (affirming CIT’s holding that Commerce may not use the petition rate to establish the dumping margin when its own investigation revealed that the petition rate was not credible). The adjusted petition rate did not and does not represent commercial reality. Commerce calculated the adjusted petition rate based on the highest dumping margin alleged in the petition. The fact that Commerce ultimately imposed dumping margins between 5.91% and 6.82% for the same products after its initial investigation shows the possession of better information and shows that the adjusted petition rate[*1324] was aberrational. Cooperating respondents’ actual dumping margins during the administrative review period — ranging from 2.58% to 10.75% — further call into question the credibility of the adjusted petition rate.

The 57.64% adjusted petition rate is more than ten times higher than the average dumping margin for cooperating respondents. This high rate is also more than five times higher than the highest rate applied to a cooperating respondent. Moreover, nothing in the record ties the adjusted petition rate to Gallant, because Gallant did not participate in the original investigation. Thus, the record shows that the 57.64% rate is unrelated to commercial reality and, thus, not a “reasonably accurate estimate” of Gallant’s actual dumping rate. Id. at 1032. Instead, the AFA rate is “punitive, aberrational, or uncorroborated,” id., and excessive in view of the cooperative respondents’ dumping rates. This court also perceives that a rate over five times the highest rate imposed on similar products is far beyond an amount sufficient to deter Gallant and other uncooperative respondents from future noncompliance.

Instead of relying on the adjusted petition rate, Commerce should have relied on more reliable “facts otherwise available” such as the representative dumping rates of similarly-sized and similarly-situated exporters in the original investigation and in the administrative review. Given that over a dozen respondents submitted timely questionnaires during the administrative review, Commerce had abundant resources from which to calculate a reasonable AFA rate. Although Commerce has discretion in choosing from a list of secondary information to support its adverse inferences, Commerce must select secondary information that has some grounding in commercial reality.

In addition, Commerce failed to corroborate the adjusted petition rate with “independent sources that are reasonably at [its] disposal.” 19 U.S.C. § 1677e(c). Commerce used a very small percentage of the mandatory respondents’ transactions as corroborative evidence even though most transactions during the period of review had significantly lower dumping margins. The record does not show that the transactions at and above the 57.64% dumping margin reflect Gallant’s commercial activity. Because Commerce did not identify any relationship between the small number of unusually high dumping transactions with Gallant’s actual rate, those transactions cannot corroborate the adjusted petition rate.

This court recognizes a distinction between the present case and both Ta Chen Stainless Steel Pipe, Inc. v. United States, 298 F.3d 1330 (Fed.Cir.2002), and PAM, S.p.A. v. United States, 582 F.3d 1336 (Fed.Cir.2009). In Ta Chen, Commerce applied a 30.95% AFA rate against Ta Chen, which did not fully cooperate with the ITC during its administrative review. 298 F.3d at 1339. In that case, the record showed that Ta Chen had made a sale with a 30.95% dumping margin during the relevant period of review. Id. Although one sale by itself does not always rise to the level of substantial evidence, the 30.95% AFA rate was reasonable because Commerce tied it to Ta Chen’s actual sales. In contrast, Commerce did not tie the AFA rate to Gallant’s actual dumping margin. Also, Ta Chen was not a corroboration case as Commerce relied on primary information — -i.e., Ta Chen’s sales data from the relevant review period — in calculating the AFA rate. Id.

[*1325] In PAM, Commerce assessed a 45.49% AFA rate against PAM because sales data from a previous administrative review showed that PAM had twenty-nine transactions with dumping margins at or above 45.49%. 582 F.3d at 1338^0. Although those transactions amounted to only 0.5% of PAM’s total U.S. sales in a previous administrative review, this court, in view of the entire record, found that the transactions were reasonably tied to PAM’s actual dumping margin. Id. at 1340. Unlike PAM, Commerce in the present case did not show that a small percentage of the mandatory respondents’ transactions represented a reasonably accurate estimate of Gallant’s actual dumping margin. Instead, the record showed a large body of reliable information suggesting the application of a much lower margin. Substantial evidence requires Commerce to show some relationship between the AFA rate and the actual dumping margin.

Although this court remands for recalculation of the AFA rate, this court finds that Commerce did not err by using transactions with broken shrimp and non-broken shrimp as corroborative evidence. The antidumping duty order covers both types of shrimp. Also, some of the Thai exporters seem to have dumped broken shrimp into the United States during the relevant review period. Likewise, Gallant could be dumping broken shrimp. Of course, if broken shrimp represent a minority of the sales, Commerce cannot base the AFA rate entirely on broken-shrimp transactions. But on this record, Commerce did not err by using some transactions with broken shrimp as corroborative evidence.

V.

Accordingly, substantial evidence does not support the unreasonably high AFA rate imposed against Gallant. This court therefore vacates and remands the Court of International Trade’s decision so that it may remand the case back to Commerce for further proceedings consistent with this opinion.

VACATED and REMANDED

COSTS

Each party shall bear its own costs.