INA Walzlager Schaeffler KG and INA Bearing Company, Inc. (collectively INA) appeal from a decision of the Court of International Trade upholding a Commerce Department ruling in an antidumping case. In the challenged ruling, Commerce determined that certain INA roller bearings are within the scope of an antidumping duty order concerning cylindrical roller bearings imported from Germany. We agree with the Court of International Trade that, under the deferential standard of review applicable in this setting, Commerce’s scope determination must be upheld.
I
In 1988, the Torrington Company, a United States manufacturer of antifriction bearings, filed a petition with the Commerce Department requesting the imposition of an-tidumping duties on imports of certain anti-friction bearings from Germany and eight other countries. In response to the petition, Commerce initiated an investigation. Commerce’s Office of Investigations prepared a “class or kind” memorandum, which concluded that the petition actually encompassed five different kinds of bearings: (1) ball bearings; (2) spherical plain bearings; (3) spherical roller bearings; (4) cylindrical roller bearings (CRBs); and (5) needle roller bearings (NRBs). The memorandum noted that the “general physical characteristics of bearings are significantly different with regard to the shape of the rolling element contained within the bearing.”
At the end of its investigation, Commerce concluded that each of the five identified classes of bearings were being sold in the United States at less than fair value. Commerce included under the CRB designation “all antifriction bearings which employ cylindrical rollers as the rolling element” and under the NRB designation “all antifriction bearings which employ needle rollers as the rolling element.”
As required by 19 U.S.C. § 1673a(a), the International Trade Commission (ITC) conducted an investigation to determine whether the importation of the antifriction bearings was materially injuring, threatening to injure, or retarding the establishment of any U.S. industry. The ITC determined that imports of CRBs from Germany were causing injury to a U.S. industry, but that imports of NRBs from Germany were not.
Following the ITC’s injury determination, Commerce issued an antidumping duty order that covered CRBs, but not NRBs, from Germany. Shortly thereafter, a German bearing manufacturer, FAG Kugelfiseher Georg Schaefer KGaA (FAG), requested that Commerce interpret the scope of the order to exclude certain engine bearings on the ground that the bearings were needle roller bearings and therefore were not within the scope of the order.
*303In conducting the FAG scope determination, Commerce concluded that “the petition and the underlying investigations demonstrate that the length-to-diameter ratio of a bearing is the key factor to distinguish a needle roller bearing from a cylindrical roller bearing.” Commerce added, however, that “neither the petition nor the underlying investigations conclusively establish a specific minimum ratio” for distinguishing between CRBs and NRBs. Commerce noted, for example, that the petition stated that a roller element length-to-diameter ratio of 2.5-to-l generally distinguishes CRBs from NRBs, while the staff report underlying the ITC’s final determination referred to a ratio of 4-to — 1 as “the appropriate cut-off.” Because of the absence of a clear line of demarcation in the administrative record, Commerce consulted the Mechanical Engineers’ Handbook (Theodore Baumeister ed., 1958) (Marks’ Handbook), a standard reference in the field, which characterized CRBs as having “[r]oller bearings with short straight rollers” and NRBs as having “rollers whose length is at least four times the diameter.” Commerce then ruled that CRBs were to be distinguished from NRBs based on the dimensions of their roller elements, and that bearings with roller elements having a length-to-diameter ratio of less than 4-to-l would be considered CRBs under the anti-dumping duty order.
Commerce applied the 4-to-l length-to-diameter test in subsequent decisions arising from the German antidumping duty order and a similar order applicable to antifriction bearings from Japan. See NTN Bearing Corp. of Am. v. United States, 905 F.Supp. 1083 (Ct. Int’l Trade 1995); Koyo Seiko Co. v. United States, 834 F.Supp. 1401 (Ct. Int’l Trade 1993), aff'd, 31 F.3d 1177 (Fed.Cir.1994) (unpublished table decision); Nippon Thompson Co. and IKO Int’l, 58 Fed.Reg. 11209 (Dep’t Commerce 1992). In two of those eases (Koyo Seiko and NTN Bearing), the importers sought judicial review, and the 4 — to—1 test was upheld in both cases.
In late 1992 INA submitted a request for a scope ruling, in which it asked Commerce to rule that all of the articles in certain INA bearing series were outside the scope of the antidumping duty order. In its request, INA urged Commerce to define NRBs according to “common industry standards” and argued that “[t]he evolution of the scope test has now resulted in a definition (the ‘4 to 1 test’) that bears no relationship whatsoever to clearly established industry standards and practices.” INA argued that in the FAG scope determination Commerce should not have relied upon Marks’ Handbook, but “should have referenced the sales catalogs of the bearing manufacturers and the standards documented by ISO [International Organization for Standardization] and DIN [Deutsch-es Instituí für Normen eV].” To show that the bearings covered by its scope request were recognized in the industry as needle roller bearings, INA included pages from its catalog identifying INA bearings according to series designations, the ISO and DIN classifications of bearings, and pages from the Torrington Company catalog listing equivalent bearings as needle roller bearings. INA also asserted that, should Commerce continue to rely on a length-to-diameter ratio for distinguishing between CRBs and NRBs, it should adopt the 2.5-to-l ratio set forth in Torrington’s petition.
Commerce rejected INA’s request and adhered to the 4-to-l length-to-diameter test established in the FAG scope determination. INA then sought review of Commerce’s ruling in the Court of International Trade, asserting that the 4-to-l test would improperly expand the scope of the antidumping duty order and that Commerce had improperly disregarded industry standards in adopting the 4 — to—1 test. Relying on its prior rulings in the Koyo Seiko and NTN Bearing cases, the Court of International Trade held that Commerce’s 4-to-l ratio test was a permissible standard for determining the scope of the antidumping duty order. Because the roller elements of INA’s bearings were not thin enough to qualify the bearings as NRBs under the 4-to-l ratio test, the court sustained Commerce’s determination that INA’s bearings were CRBs and therefore fell within the scope of the antidumping duty order. The Court of International Trade subsequently reached the same conclusion in reviewing the final results of Commerce’s third annual administrative review of the anti-*304dumping duty order. INA then took this appeal from both rulings of the Court of International Trade, challenging Commerce’s use of the 4-to-l test for distinguishing CRBs from NRBs.
II
Antidumping duties are assessed when Commerce determines that a class or kind of merchandise is being, or is likely to be, sold in the United States at less than its fair value and the ITC determines that the importation of such merchandise is causing or threatening to cause material injury to a United States industry, or is materially retarding the establishment of an industry in the United States. See 19 U.S.C. § 1673. In response to an application by an interested party to determine whether a particular product is within the scope of an antidump-ing duty order, see 19 C.F.R. § 353.29(b), Commerce considers the following:
(1) The descriptions of the merchandise contained in the petition, the initial investigation, and the determinations of the Secretary and the Commission.
(2) When the above criteria are not dispos-itive, the Secretary will further consider:
(i) The physical characteristics of the product;
(ii) The expectations of the ultimate purchasers;
(in) The ultimate use of the product; and
(iv) The channels of trade.
19 C.F.R. § 353.29®.
It is undisputed that the bearings at issue in INA’s scope request have a length-to-diameter ratio of less than 4-to-l and thus constitute CRBs under Commerce’s test. The only issue before us is whether the line Commerce has drawn to distinguish CRBs from NRBs is unsupported by substantial evidence or is contrary to law. See 19 U.S.C. § 1516a(b)(l).
Commerce’s definition of CRBs for purposes of the antidumping duty order has two components. First, Commerce has determined that the distinction between CRBs and NRBs turns on the length-to-diameter ratio of their roller elements. Second, Commerce has selected a length-to-diameter ratio of 4-to-1 as the fine at which the distinction is to be drawn. There is clearly substantial evidence in the administrative record to support Commerce’s conclusion as to the first component — that the distinction between CRBs and NRBs in the antidumping duty order is based on the ratio of the length of the roller element to its diameter. It is a closer question whether there is substantial evidence to support Commerce’s conclusion as to the second component — that 4-to-l is the proper length-to-diameter ratio to distinguish between CRBs and NRBs. Nonetheless, in light of Commerce’s broad authority to interpret the scope of its anti-dumping duty orders, we sustain its selection of the 4-to-l ratio.
A
INA contends that in interpreting the anti-dumping duty order, Commerce erred by looking beyond the definitions of CRBs and NRBs contained in the petition, the original investigation notices, the Commerce and ITC determinations, and the order itself. Those definitions, according to INA, were meant to reflect the ordinary meaning of the terms in the industry and do not support Commerce’s selection of the 4-to-l length-to-diameter ratio as the dividing line between CRBs and NRBs.
Contrary to INA’s contention, the administrative record leading to the issuance of the antidumping duty order does not provide much guidance as to the distinction between CRBs and NRBs. The petition characterizes NRBs as “a type of cylindrical bearing distinguished by a comparatively small diameter and a high ratio of length to diameter.” The notices and the order are likewise imprecise in that respect. Both characterize “needle bearings” as bearings having “needle rollers,” and “cylindrical bearings” as bearings having “cylindrical rollers,” without further defining the terms “needle rollers” and “cylindrical rollers.”
While the petition, the notices, and the antidumping duty order do not draw a clear line of distinction between the two types of bearings at issue in this case, they do have this important feature in common: they *305teach that the classification of a bearing as a CRB or an NRB is determined by the physical characteristics of its rolling element. The materials that Commerce and the ITC consulted in the course of their investigations likewise focused on the physical characteristics of the roller elements in distinguishing between CRBs and NRBs. Commerce’s “class or kind” memorandum drew distinctions among the various types of roller bearings “based on the type of rolling element contained in each type of bearing.” The staff report that accompanied the ITC’s “material injury” determination similarly focused on the length-to-diameter ratio of the roller elements as the critical feature distinguishing CRBs from NRBs. And the reports that preceded the ITC’s “material injury” determination also identified the shape of the roller element as the distinguishing feature of the two types of bearings. See International Trade Comm’n, USITC Pub. No. 1797, Competitive Assessment of the U.S. Ball and Roller Bearing Industry 11 (Jan.1986) (“Needle roller bearings are a special type of cylindrical bearing, distinguished by a comparatively small diameter and high ratio of length to diameter.”); International Trade Comm’n, USITC Pub. No. 2083, Antifriction Bearings (Other Than Tapered Roller Bearings) And Parts Thereof From the Federal Republic of Germany, France, Italy, Japan, Romania, Singapore, Sweden, Thailand, and the United Kingdom A — 4 (May 1988) (preliminary determination) (“Roller bearings are classified according to the shape of the roller used.”).
Because the materials in the administrative record repeatedly focused on the physical properties of the roller element as the basis for distinguishing CRBs from NRBs, we conclude that Commerce’s decision to distinguish between NRBs and CRBs on the basis of the length-to-diameter ratio of the roller element is supported by substantial evidence. We therefore reject INA’s argument that Commerce should not have based its scope determination on the physical properties of the roller elements, but should have looked to the ISO and DIN classifications or the designations used by INA and Torrington in their catalogs.
B
While the materials in the administrative record amply support the use of the length-to-diameter ratio to categorize cylindrical bearings, those materials are not nearly as helpful in identifying the particular ratio that should be used to draw the line between cylindrical and needle bearings. The original Tomngton petition characterized NRBs as having “[generally a ratio of length to diameter of 2.5 to one.” The staff report that was prepared during the course of the ITC investigation pointed to a different ratio, stating that needle bearings “often” have a roller element with a length “at least four times greater than its diameter.” The synopsis of a report prepared by Commerce’s Office of Industrial Resources Administration (OIRA), which Commerce consulted in the course of its investigation, adopted that same ratio, although more definitively: it stated that “[njeedle roller bearings have rolling elements with lengths at least four times greater than diameter.”
In its final determination, the ITC cited the Harmonized Tariff Schedule of the United States (HTSUS), which distinguishes between needle roller bearings and other cylindrical roller bearings by reference to the structure of their roller elements. Under the HTSUS, roller bearings with cylindrical rollers not exceeding five millimeters and “having a length which is at least three times the diameter” are classified as needle roller bearings, and all other bearings with cylindrical rollers are classified as cylindrical roller bearings. The ITC’s reference to the tariff schedule would seem to provide support for adoption of a 3-to-l length-to-diameter test for distinguishing between CRBs and NRBs, but INA does not contend that the HTSUS classification resolves the definitional issue. In fact, INA argues that “tariff classifications are not dispositive of scope issues,” and that the ITC’s reference to the tariff schedules does no more than indicate that the ITC did not mean to adopt a 4-to-l length-to-diameter ratio as the test for distinguishing CRBs from NRBs.
In the scope ruling under review here, Commerce followed its prior, more de*306tailed scope ruling in the FAG case, in which Commerce first settled on the 4-to-l length-to-diameter ratio as the proper test for distinguishing between CRBs and NRBs. In the FAG scope ruling, Commerce explained that in choosing among competing length-to-diameter ratios it had consulted Marks’ Handbook, which characterizes NRBs as having a length “at least four times the diameter.” INA asserts that Marks’ Handbook does not reflect industry standards for distinguishing between CRBs and NRBs, and that it was therefore error for Commerce to rely on that publication. INA’s argument, however, ignores the first step in Commerce’s analysis. Once Commerce decided that the antidump-ing duty order distinguishes NRBs from CRBs on the basis of the length-to-diameter ratio of their roller elements, it was no longer relevant that the manufacturers’ catalogs and the ISO and DIN standards classified NRBs in a different manner. Commerce’s task at that point was simply to select the most suitable ratio of length to diameter. With respect to that issue, the 4-to-l ratio set forth in Marks’ Handbook supported the definition suggested by the ITC staff report and adopted in the OIRA synopsis. In light of the conflict among the sources to which Commerce had reference and the absence of any dispositive ratio that Commerce was legally compelled to adopt, we conclude that it was permissible for Commerce to select the 4-to-l length-to-diameter ratio as the dividing line between CRBs and NRBs.
C
In addition to challenging Commerce’s scope ruling on substantial evidence grounds, INA points to several specific flaws in the ruling and contends that those flaws undermine the validity of Commerce’s decision. While INA’s criticisms of the ruling are well founded in several respects, the flaws in the decision are not fatal.
One of INA’s arguments is that Commerce’s reliance on a long-outdated edition of Marks’ Handbook, rather than the more current edition, undermines the validity of Commerce’s selection of the 4-to-l test. The current edition of Marks’ Handbook, like the edition on which Commerce relied, defines NRBs by reference to the 4-to-l ratio, but the current edition also refers to CRBs as having rolling elements “with approximate length-diameter ratio ranging from 1:1 to 1:3,” a definitional provision not found in the earlier edition. The new material, INA argues, points away from the 4-to-l ratio test that Commerce adopted.
Commerce’s use of an outdated edition of a reference work on which it placed significant reliance certainly seems to reflect careless staff work. It does not, however, require reversal, because the error was not prejudicial in the circumstances of this case. INA acknowledges that, for purposes of the antidumping duty order, any cylindrical roller bearing that is not an NRB is necessarily a CRB. Because NRBs were specifically excluded from the coverage of the antidumping duty order, it was proper for Commerce to focus on the definition of NRBs in determining the scope of the order. While the definition of CRBs changed between the sixth and ninth editions of Marks’ Handbook, the definition of NRBs remained constant. It was therefore permissible for Commerce to rely on the Marks’ Handbook definition of NRBs to distinguish between the two classes of bearings, even though the edition of Marks’ Handbook that Commerce consulted was long out-of-date.
On the same theme, INA notes that the ITC staff report, which used the 4-to-l test in referring to NRBs, also stated that CRBs have rollers that are “approximately equal in length and diameter.” That reference, INA argues, undermines whatever support the staff report provides for the 4-to-l ratio as the dividing line between CRBs and NRBs. The staff report’s reference to CRBs, however, was sufficiently general that it was not necessarily inconsistent with the more specific 4-to-l ratio that the report used in defining NRBs. Moreover, it is reasonable to conclude that it was the definition of NRBs that informed the ITC’s decision that the importation of NRBs was not causing or threatening material injury to any U.S. industry. In addressing the scope of the antidumping duty order, Commerce therefore permissibly focused on the definition of NRBs that the ITC had before it when making its “material injury” determination.
*307Finally, pointing to Commerce’s statement in the INA scope ruling that the ITC “relied upon the 4 to 1 ratio to distinguish cylindrical and needle roller bearings,” INA contends that Commerce’s characterization of the ITC proceeding was incorrect and that the scope ruling must be overturned for that reason. We agree with INA that Commerce’s statement that the ITC “relied on” the 4-to-l ratio in maMng its material injury determination and thus required Commerce to adopt that test is at least an overstatement. Nonetheless, that mischaracterization is harmless, because Commerce based its ruling in this case primarily on the analysis it had employed in the FAG scope ruling. There, as we have noted, Commerce permissibly concluded that the definitions of CRBs and NRBs for purposes of the antidumping duty order turned on the length-to-diameter ratio of the'roller elements in the bearings, and that the 4r-to-l ratio was the best line of demarcation between the two. That analysis was sufficient to reject INA’s reliance on its catalog and the ISO and DIN bearing designations. Commerce’s remarks about the extent to which the ITC embraced the 4-to-l standard were therefore not critical to its decision in this case and accordingly do not require that the scope determination be overturned.
In sum, we hold that Commerce’s determination in this case fell within the agency’s broad authority to interpret its own anti-dumping duty orders. See Ericsson GE Mobile Communications, Inc. v. United States, 60 F.3d 778, 782 (Fed.Cir.1995) (Commerce “enjoys substantial freedom to interpret and clarify its antidumping duty orders”). We therefore affirm the decision of the Court of International Trade upholding Commerce’s determination that INA’s subject bearings are within the scope of the antidumping duty order on antifriction bearings imported from Germany.
Each party shall bear its own costs for this appeal.
AFFIRMED.