OPINION
GOLDBERG, Senior Judge:Plaintiff Ad Hoc Shrimp Trade Action Committee (“Ad Hoc Shrimp”) is a domestic association of producers and processors of warmwater shrimp. In this action, Ad Hoc Shrimp contests certain aspects of the administrative determination issued by the International Trade Administration of the United States Department of Commerce (“Commerce”) in the second administrative review of the antidumping duty order covering certain frozen warm-water shrimp from Ecuador. See Certain Frozen Warmwater Shrimp from Ecuador: Final Results and Partial Rescission of Antidumping Duty Administrative Reviews, 73 Fed. Reg. 39,945 (Dep’t Commerce July 11, 2008) {“Final Results”). Ad Hoc Shrimp alleges that Commerce erroneously accepted the raw material cost information for shrimp products reported by Defendant-Intervenor Oeeanlnvest S.A. (“Oeeanlnvest”). Accordingly, Ad Hoc Shrimp claims that Commerce’s Final Results are unsupported by substantial evidence on the record and otherwise not in accordance with the law. Commerce supports its determination to rely upon Oceanlnvest’s reported costs in the Final Results because those costs reflect the actual costs associated with the production of shrimp products and because Commerce did not depart from past practice. For the reasons that follow, the court sustains Commerce’s determinations and denies Plaintiff’s motion for judgment on the agency record to remand the Final Results to Commerce.
I.
JURISDICTION AND STANDARD OF REVIEW
The Court has jurisdiction over this matter pursuant to section 516a(a)(2)(B)(iii) of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2006), and 28 U.S.C. § 1581(c) (2006).
For administrative reviews of antidumping orders, the Court sustains Commerce’s determinations, findings, or conclusions unless they are “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i). Substantial evidence is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Universal Camera Corp. v. NLRB, 340 U.S. 474, 477 (1951). When a party alleges that Commerce’s action is not supported by substantial evidence, the Court assesses whether the agency action is “unreasonable” given the *1652record as a whole. Nippon Steel Corp v. United States, 458 F.3d 1345, 1350-51 (Fed. Cir. 2006). “[T]he possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence.” Consolo v. Fed. Mar. Comm’n, 383 U.S. 607, 620 (1966) (citing NLRB v. Nevada Consol. Copper Corp., 316 U.S. 105, 106 (1942)). The Court need only find evidence “which could reasonably lead” to the conclusion drawn by Commerce, thus making it a “rational decision.” Matsushita Elec. Indus. Co., v. United States, 750 F.2d 927, 933 (Fed. Cir. 1984).
II.
STATEMENT OF THE FACTS
Oceanlnvest produces finished, i.e. “value-added”, shrimp products from raw shrimp delivered by farmers to Oceanlnvest’s processing facility in Ecuador. The production process includes classifying shrimp based upon its size. The size of shrimp, referred to as the count size, is expressed in terms of the number of individual shrimp contained in a unit of weight. For example, a count size of 51/60 headless shrimp indicates one pound of shrimp that contains 51 to 60 individual headless shrimp. Individual shrimp may vary in size within the count size classification for a value-added shrimp product. Oceanlnvest pays a higher price for larger shrimp. For instance, Oceanlnvest pays more for 51/60 count size than 61/70 count size because the latter contains smaller shrimp.
In February 2007, at Ad Hoc Shrimp’s request, Commerce initiated a sales-below-cost investigation against Oceanlnvest. The investigation focused on Oceanlnvest’s sales of frozen warmwater shrimp in the United States during the period of review (“POR”) from February 1, 2006 to January 31, 2007. Over the course of the administrative review, Commerce sent Oceanlnvest three sets of supplemental questionnaires inquiring into its cost of production (“COP”) reporting. On March 6, 2008, Commerce published the preliminary results of its administrative review. Certain Frozen Warmwater Shrimp from Ecuador, 73 Fed. Reg. 12,115 (Dep’t Commerce Mar. 6, 2008) (preliminary results). In its analysis, Commerce utilized Oceanlnvest’s reported costs of raw material inputs.
Ad Hoc Shrimp filed a brief contending that Commerce should reject Oceanlnvest’s reported costs of raw material inputs as distor-tive. After evaluating the information and explanations provided by Oceanlnvest, Commerce disagreed with Ad Hoc Shrimp and accepted Oceanlnvest’s reported costs in the Final Results. Decision Memorandum, A-331-802, ARP 06-07, Admin. R. Pub. Doc 165 (July 3, 2008) available at http://ia.ita.doc.gov/frn/ summary/ECUADOR *1653ZE8-15830-l.pdf (last visited October 14, 2009) (“Decision Mem”). Commerce determined that the reported-cost information was consistent with Oceanlnvest’s normal accounting records and reasonably reflected the costs associated with the production and sale of the merchandise. Id. at 11.
Ad Hoc Shrimp then filed this action against Commerce under 28 U.S.C. § 1581(c). This Court allowed Oceanlnvest to intervene.
III.
DISCUSSION
Ad Hoc Shrimp raises two arguments to support its claim that the Final Results are unsupported by substantial evidence on the record or otherwise not in accordance with law. First, it alleges that Commerce unreasonably accepted Oceanlnvest’s reported raw material input costs for value-added products despite the fact that those costs are unreliable and reflect a physical impossibility. Second, Ad Hoc Shrimp argues that Commerce’s acceptance of Oceanlnvest’s [ ] raw material costs for virtually identical value-added products is an unexplained and unsupported departure from Commerce’s past practice. The court addresses each argument in turn.
A. Commerce’s determination that Oceanlnvest’s reported raw material cost information reasonably reflects its actual production costs is supported by substantial evidence and otherwise in accordance with the law.
When considering the imposition of an antidumping duty, cost of production:
“shall normally be calculated based upon the records of the exporter or producer of the merchandise, if such records are kept in accordance with the generally accepted accounting principles of the exporting country... and reasonably reflect the costs associated with the production and sale of the merchandise.”
19 U.S.C. § 1677b(f)(l)(A). Ad Hoc Shrimp does not dispute whether Oceanlnvest’s records were kept in accordance with Ecuadorian generally accepted accounting principles (“GAAP”). Rather, Ad Hoc Shrimp claims that Commerce’s acceptance of Oceanlnvest’s reported raw material costs was not reasonable under section 773(f)(1)(A) of the Tariff Act, 19 U.S.C. § 1677b(f)(l)(A) because those costs reflected a production process that was physically impossible and unreliable. Specifically, Ad Hoc Shrimp argues that Oceanlnvest’s explanation of *1654its cost methodology is facially implausible because it [ ]1 Ad Hoc Shrimp points to record evidence for one shrimp product, as identified by the product’s control number (CONNUM), that suggests that the finished shrimp product produced does not appear attainable based on the raw material shrimp size used. See Decision Mem. at 8. However, Ad Hoc Shrimp fails to take into account several important factors underlying Oceanlnvest’s production process and accounting system.
First, Oceanlnvest explained that a mix of raw shrimp inputs, including smaller input size shrimp, can be used to produce a larger peeled product. For example, a combination of 51/60 count size and 61/70 count size raw shrimp can be used to produce a 51/60 count size peeled shrimp product. See Letter from Cameron & Hornbostel LLP to U.S. Department of Commerce, Case No. A-331-802, Admin R. Pub. Doc. 152, Non-Pub. Doc. 50, at 1-2 (Mar. 11, 2008) (third supplemental section D response) (“Third Supplemental Section D Response”).2 The resulting mix would be classified as a 51/60 finished product despite the presence of individual shrimp of varying count size.3 It is the total number of shrimp in the finished product, not the size of each individual shrimp, that determines the marked count size of the product.
The record evidence demonstrates that Oceanlnvest used different mixes of raw input shrimp count sizes to produce finished products that satisfied its customers’ size specifications. See, e.g., Admin. R. Non-Pub. Doc. 34 (first supplemental section D response). During its investigation, Commerce reviewed Oceanlnvest’s inventory tracking system and found that Oceanlnvest tracks the actual mix of shrimp input sizes that are used to produce each peeled product. Decision Mem. at 10. This system tracked both the input shrimp size and cost on an actual, as invoiced basis, to ensure that the final recorded costs accurately reflected the prices paid for the inputs. Id. at 10-11.
Moreover, Ad Hoc Shrimp’s “disappearing shrimp” argument is without merit. Ad Hoc Shrimp claims that Oceanlnvest failed to *1655explain how shrimp allegedly disappeared in the production process.4 Thus, the total number of shrimp reported by Oceanlnvest, according to Ad Hoc Shrimp, does not reasonably reflect the costs associated with its production. However, Oceanlnvest explained that it was reporting the equivalent yield of shrimp after the production process. When the shell is peeled from the shrimp, the difference between the weight of the shrimp before and after peeling is referred to as the “yield.” The reported “yield” for the control number at issue was [ ] which meant that the total weight after peeling was [ ] of the total weight before peeling. In other words, Oceanlnvest was stating that, at the end of the production process, it had the equivalent yield of [ ] individual shrimp, not that it had lost individual shrimp during the course of the production.5
Ad Hoc Shrimp also misinterprets the ramifications of the “purchasing strategy” disclosed by Oceanlnvest. Oceanlnvest explained to Commerce that, from time to time, it employs a purchasing strategy whereby it [ ] Third Supplemental Section D Response at 1, 3. Under this purchasing strategy, Oceanlnvest will [ ] See Cost of Production and Constructed Value Calculation Adjustments for the Final Results, A-331-802, ARP 06-07, Admin. R. Pub. Doc. 167, Non-Pub. Doc. 59 (July 3, 2008) (“COP Mem., Admin. R. Non-Pub. Doc. 59”). In other words, Oceanlnvest will [ ] Oceanlnvest then classifies and records the purchased shrimp [ ] and they pay the supplier [ ] price for the shrimp. Id. For the particular control number at issue, Oceanlnvest explained that it had engaged in this purchasing strategy [ ] Third Supplemental Section D Response at 3-4. Ad Hoc Shrimp does not assert that this purchasing strategy violates any statute or regulation governing the calculation of Oceanlnvest’s raw shrimp costs. Instead, Ad Hoc Shrimp argues that the conduct of these transactions renders Oceanlnvest’s reporting inaccurate because [ ] Ad Hoc Shrimp argues that this purchasing practice calls into question the rest of Oceanlnvest’s reported costs.
*1656While Ad Hoc Shrimp asserts that Oceanlnvest [ ], Commerce found that the costs are not distortive. Pursuant to the Tariff Act, the producer must report to Commerce the actual price paid for raw shrimp, 19.U.S.C. §§ 1677b(e) and (f), which Commerce found had occurred in this case. Decision Mem. at 11. In addition, Commerce noted that this purchasing practice is infrequent and represents only a small percentage of Ocean-lnvest’s overall raw material purchases.6 Id. When it did happen, Oceanlnvest recorded in its accounting system [ ] COP Mem., Admin. R. Non-Pub. Doc. 59. Oceanlnvest calculated the actual invoice cost of the raw shrimp in their normal books and records which they used to compute their reported costs to Commerce. Id. Commerce thereby determined that the costs captured in the reported costs reflected Oceanlnvest’s actual costs incurred for its raw material shrimp inputs, Decision Mem. at 11, [ ]
Simply because Ad Hoc Shrimp argues that this “purchasing strategy” is not a reasonable explanation for Oceanlnvest’s reported information does not prevent Commerce’s determination from being supported by substantial evidence. See Catfish Farmers of America v. United States, Slip Op. 09-96, 2009 WL 2921300 (CIT Sep. 14, 2009) (“The administrative record for an antidumping duty administrative review may support two or more reasonable, though inconsistent, determinations on a given issue.”) As part of its investigative process, Commerce specifically requested a deeper explanation of the reported costs for the control number at issue. Commerce reviewed how Oce-anlnvest reflected its payments to the farmers in its raw material inventory system. COP Mem., Admin. R. Non-Pub. Doc. 59. It then reviewed how the costs in that system flowed in the calculations contained in the questionnaire responses on COP reporting that Oce-anlnvest filed with Commerce. Id. Commerce analyzed Oceanlnvest’s reported costs taking into consideration Oceanlnvest’s inventory tracking system, cost methodology, and questionnaire responses explaining this purchasing strategy. Commerce thereby determined that the infrequent use of this purchasing practice did not invalidate the accuracy of Oceanlnvest’s material costs. Decision Mem. at 10-11. Based on the record evidence, Commerce reasonably concluded that *1657this purchasing strategy, in light of Oceanlnvest’s reported information and supplemental responses, was a reasonable explanation for Oceanlnvest’s reported costs.
Ad Hoc Shrimp also claims that Commerce’s determination in the Final Results is not reasonable under section 773(f)(1)(A) of the Tariff Act, 19 U.S.C. § 1677b(f)(l)(A) because agency practice requires accurate product-specific costs in addition to accurate aggregate costs. Pl.’s Br. 10 (citing Certain Preserved Mushroom from Indonesia, 63 Fed. Reg. 72,268, 72,276 (Dep’t Commerce Dec. 31, 1998) (“The fact that the inaccurate standards for each major cost element add up to a total that is closer to the actual total costs does not support the claim that individual standard costs are reliable.”)). This argument reiterates Ad Hoc Shrimp’s previous argument regarding [ ] because Ad Hoc Shrimp frames the issue not as whether Oceanlnvest’s costs reported were accurate, but instead focuses on the reported count size information. It alleges Commerce unreasonably accepted product-specific raw material count size information [ ]
However, the record indicates that Commerce did not accept Oce-anlnvest’s reported costs only because they were reported accurately on an aggregate basis. Oceanlnvest followed Commerce’s normal practice and reported the model-specific average shrimp costs incurred during the period of review for each category of products. See Decision Mem. at 10. Moreover, Oceanlnvest [ ] Commerce understood how Oceanlnvest recorded raw material costs and sizes in its accounting system. Commerce concluded that it is reasonable to accept Ocean-lnvest’s explanation that different raw shrimp inputs could produce “value-added products of the same finished count size although a POR average cost is used for each raw shrimp input in calculating production costs.” Id.
In summary, Commerce’s determination that Oceanlnvest’s reported raw material cost information for value-added products reasonably reflects its actual production costs is supported by substantial evidence and is in accordance with the law. Pursuant to section 773(f)(1)(A) of the Tariff Act, 19.U.S.C. § 1677b(f)(l)(A), Commerce relied on Oceanlnvest’s recorded information because it was kept in accordance with the home country GAAP. This Court has consistently upheld Commerce’s reliance on a company’s costs as recorded in its financial statements “as long as those statements were prepared in accordance with the home country’s GAAP and do not significantly distort the firm’s actual costs.” Solvay Solexis S.P.A. v. U.S, 628 F.Supp.2d 1375, 1379 (CIT 2009); see also Cinsa, S.A. de C.V. v. *1658United States, 21 CIT 341, 343, 966 F. Supp. 1230, 1235 (1997); FAG U.K. Ltd. V. United States, 20 CIT 1277, 1290, 945 F. Supp. 260, 271 (1996). Ad Hoc Shrimp fails to demonstrate that the reported costs are significantly distorted. Commerce gained an understanding of how Oceanlnvest uses mixtures of different count sizes of raw material to produce the same finished products. It also found that Ocean-Invest’s inventory system tracks the specific mix of actual shrimp inputs for each finished product. Moreover, Commerce understood that, in rare instances, Oceanlnvest used raw material that it recorded as [ ] and found that Oceanlnvest reported the actual prices it paid. Based upon the record evidence, Commerce reasonably determined that Oceanlnvest reported the actual raw material costs needed to produce the value-added products and that these costs were not distortive.
B. Commerce’s acceptance of Oceanlnvest’s different raw material costs for similar value-added products is not inconsistent with the Agency’s practice.
Ad Hoc Shrimp argues that Commerce erroneously accepted [ ] raw material costs for virtually identical finished shrimp products in violation of Commerce’s settled practice. Specifically, Ad Hoc Shrimp indicates that, for two sets of control numbers, Ocean-lnvest reported different raw material count sizes [ ] for the same peeled shrimp products differing only in terms of container weight or presentation, two physical characteristics that Ad Hoc Shrimp alleges have no bearing on raw material costs. Ad Hoc Shrimp asserts that, under established practice, identical products must have identical costs reported for them.
Under Commerce’s methodology, Oceanlnvest must calculate its COP on a control number-specific basis. The control number identifies a shrimp product by the physical characteristics that Commerce determines can have a material effect on product prices and costs.7 In the shrimp investigations, Commerce identified 14 distinct physical characteristics, including container weight and presentation, that could have such an effect. Thus, under Commerce’s methodology, if the container weight or presentation differs, the control number differs. Each product that has a different control number is a different product for cost calculation purposes, no matter how physically similar those products may be as a practical matter.
The products at issue do not “share the same physical characteristics” as defined by Commerce because they differ in container weight or presentation. Accordingly, they are not identical products and do *1659not require identical material costs. The record indicates that different mixes of shrimp count sizes as inputs can produce the same value-added product that differ only in container sizes or presentations. See, e.g., Admin. R. Non-Pub. Doc. 42 (second supplemental section D response). Therefore, Commerce’s methodology permits different raw material count size and cost information for shrimp products differing only in their container size or their presentation.
Moreover, the court is not convinced that Commerce’s acceptance of Oceanlnvest’s different raw material costs for certain shrimp products differing only in container weight or presentation deviates from past practice. Ad Hoc Shrimp argues that Commerce deviated from its past practice by accepting Oceanlnvest’s reported raw material costs in the second administrative review even though the costs suffer from the same deficiency as the costs that Commerce rejected in the first administrative review. See Certain Frozen Warmwater Shrimp from Ecuador: Final Results of Antidumping Duty Administrative Review, 72 FR 52070 (Dep’t Commerce Sept. 12, 2007) (first administrative review). However, Commerce’s refusal to accept Oceanln-vest’s raw material costs in the first administrative review concerned separate issues. In the first administrative review, Commerce determined that Oceanlnvest had made errors in calculating its raw shrimp costs which Oceanlnvest then corrected. See Decision Memorandum, A-331-802, AR 04-06 (Sep. 5 2007), available at 2007 WL 2773557 (issues and decision memorandum to first administrative review). Oceanlnvest had reported its costs based on finished shrimp count size rather than input shrimp count size. Id. at cmt. 6. It also incorrectly reported the last purchase price in the month in which a shrimp product was actually produced rather than calculate the weighted average cost of raw material that it purchased in the entire POR. Id. In other words, Commerce found that Oceanlnvest’s reported costs were distorted for reasons separate from Oceanlnvest’s COP reporting in the second administrative review.
In both administrative reviews, Commerce followed its practice of requiring a respondent to report costs on a product-specific basis and also of relying upon a respondent’s normal books and records. Moreover, in the first administrative review, Commerce stated that a respondent must calculate its raw shrimp costs based on the physical characteristics as defined by Commerce. Id. (emphasis added). Under Commerce’s methodology, if the container weight or presentation differs, the control number differs. Thus, in both reviews, Commerce was to treat container weight and presentation as separate physical characteristics in the control number. Therefore, Commerce’s acceptance of Oceanlnvest’s reported costs in the Final Results was reasonable because those costs comported with Commerce’s normal prac*1660tice, which is the statutorily preferred methodology pursuant to section 773(f)(1)(A) of the Tariff Act, 19.U.S.C. § 1677b(f)(l)(A).
IV.
CONCLUSION
Commerce’s determination that Oceanlnvest’s reported raw material cost information for value-added products reasonably reflects its actual production costs is supported by substantial evidence and otherwise in accordance with the law. Commerce’s acceptance of Oce-anlnvest’s different raw material costs for similar finished shrimp products does not depart from agency practice. For the foregoing reasons, the Court sustains Commerce’s final determination and denies Ad Hoc Shrimp’s motion for judgment on the agency record.
Ad hoc explains that removing the shell necessarily reduces, not increase, the weight of the raw material. Therefore, Ad Hoc Shrimp asserts that even if all of the shell-on raw material Oceanlnvest listed as being [ ] were in fact [ ], it would still be impossible to produce a peeled product weighing [ ] per piece as Oceanlnvest reported.
Further references to Oceanlnvest’s supplemental questionnaire responses are cited to the administrative record.
Oceanlnvest elaborates that a processor could mix 51 count size and 65 count size shrimp on a 50/50 basis and produce a one pound box containing 59 shrimp. In an extreme example, a shrimp processor could use 90% 61 count peeled shrimp and 10% 51 count peeled shrimp and still produce an average count size of less than 60, which falls within the specified 51/60 range.
This claim focuses on Oceanlnvest’s questionnaire response that suggests that the company started the production process with [ ] shrimp but ended up with only [ ] shrimp. Third Supplemental Section D Response at 2-3.
As Oceanlnvest explains, [ ] Commerce did not understand Oceanlnvest to say that [ ] individual shrimp before peeling are literally equal to [ ] after peeling.
Commerce explained that the record evidence demonstrates that this purchasing practice affected [ ]. Admin. R. Non-Pub. Doc. 34 at Ex. SD-11; Admin. R. Non-Pub. Doc.42 at Ex. 2SD-6. These [ ] represent only [ ] percent of the[ ] individual reported peeled control numbers, and further, peeled products as a whole account for only [ ] percent of total shrimp production for Ocean Invest during the period of review. Def.’s Br. at 9.
Different physical characteristics include form (raw or cooked), head status (head-on or headless), count size, and shell status (shell-on or peeled).