FRIEDMAN, Circuit Judge.

This is an appeal from the judgment of the United States Court of International Trade entered on remand from our decision in Schott Optical Glass, Inc. v. United States, 750 F.2d 62 (Fed.Cir.1984) (Schott II). On remand, the appellant was given the opportunity to demonstrate that a prior Court of Customs and Patent Appeals decision construing the term “optical glass” in a customs classification was clearly erroneous. The Court of International Trade held that appellant had failed so to demonstrate, and sustained the Customs Service’s classification of appellant’s imported glass as “optical glass.” Schott Optical Glass, Inc. v. United States, 678 F.Supp. 882 (1987). We affirm.

*867 I

The background facts were set forth in our prior opinion in Schott II:

The merchandise consists of seven types of filter glass used in optical instruments such as spectrometers, spec-trophotometers and solar filter simulators. Two of the filters — one nearly colorless and the other colorless — transmit visible light while absorbing specific wavelengths in the ultraviolet or infrared spectrum. The remaining five types are dark glass which absorb most of the visible light, transmitting either the ultraviolet or infrared light.

The Customs Service classified the importations as “other optical glass” under item 540.67 of the Tariff Schedules of the United States (TSUS). The appellant (Schott) contends that six of the seven should have been classified as “colored or special glass” under item 542.92 and the remaining type as “ordinary glass” under item 542.42.

There was an earlier case between the same parties in which the courts upheld the Customs Service's classification as “optical glass” of colored filter glass imported by the appellant. In Schott Optical Glass, Inc. v. United States, 468 F.Supp. 1318 (Cust.Ct.), aff'd, 612 F.2d 1288 (CCPA 1979) (Schott I), the Customs Court, on the basis of prior decisions, ruled that glass classified as optical glass under item 540.67 TSUS is presumed to be (a) very high quality, (b) used for optical instruments, and (c) capable of performing an optical function, 468 F.Supp. at 1322. It held that Schott failed to sustain its burden of proving that its glass did not come within the foregoing standards. The Court of Customs and Patent Appeals affirmed, holding that the common meaning of optical glass (which governed its tariff classification) does not include the additional requirement (which Schott urged it did) that the glass have a specifically controlled refractive index and dispersion. 612 F.2d at 1286.

750 F.2d at 63.

In Schott II, we reviewed a Court of International Trade decision that refused to allow Schott to introduce evidence to show that the Court of Customs and Patent Appeals’ interpretation of the term “optical glass” in Schott I was clearly erroneous. That court had ruled that it was bound by the prior determination in Schott I and that relitigation by Schott of either the meaning of “optical glass” or the classification of its filters as within the “other optical glass” category was barred by stare decisis. Schott Optical Glass, Inc. v. United States, 587 F.Supp. 69, 71 (1984).

We reversed, ruling that the court had incorrectly applied res judicata principles to the different issue of whether Schott should be afforded an opportunity to demonstrate that stare decisis was inapplicable because the prior decision was clearly erroneous. We pointed out that United States v. Stone & Downing Co., 274 U.S. 225, 235-36, 47 S.Ct. 616, 618-19, 71 L.Ed. 1013 (1927), held that in customs classification cases a determination of law or fact with respect to one importation is not res judi-cata with respect to another importation of the same merchandise by the same parties. We remanded to permit Schott to submit the additional evidence it believed would undermine Schott I. 750 F.2d at 65.

At an eleven-day trial on the remand, Schott presented the testimony of six witnesses and introduced numerous excerpts from dictionaries, scientific authorities and tariff literature to support its contention that colored filter glass is not “optical glass.” Schott contended that the definition of “optical glass” announced in Schott I must include two additional factors: the glass must be (1) colorless and (2) controlled for refractive index and dispersion.

After considering the evidence, the Court of International Trade held that Schott had failed to demonstrate that the determination in Schott I was clearly erroneous. The court further held that in any event, Schott had failed to overcome the presumption of correctness of Custom’s classification of its colored filter glass as optical glass.

Schott’s burden on remand was to demonstrate that the decision in Schott I was *868 clearly erroneous. In Schott II we specifically declined to consider whether our predecessor court’s prior decision was undermined by Schott’s “new” evidence. As we noted:

Schott, in effect, made an extensive offer of proof of what the new evidence would show. We cannot say whether that evidence, when added to the evidence in the prior case and subjected to cross-examination when testimonial, would cause either the Court of International Trade or this court on appeal to conclude that the prior decision was clearly erroneous. Considering all the circumstances of this case, however, we conclude that Schott should have been allowed to offer the additional evidence it believed would undermine the prior decision.

In so ruling, we intimate no opinion on what effect, if any, that evidence would have in evaluating the correctness of Schott I.

750 F.2d at 65.

On remand, Schott was permitted to introduce all the evidence it believed would show that Schott I was clearly erroneous. The Court of International Trade considered the legislative history, case law, dictionary definitions and scientific authorities that Schott cited, but concluded that they did not warrant rejection of the Schott I standard. The court noted that some of the definitions seemed inconsistent.

The testimony of Schott’s own witnesses undercut its position that colored filter glass is not optical glass. Dr. Marker, a Schott employee in charge of research, quality control and technical services, testified that black infrared transmitting glass could be optical glass. 678 F.Supp. at 890-91. He further testified that the colored filter glass at issue meets the American Society for Testing and Materials (ASTM) standard definition for optical glass. Id. at 891. In addition, Schott’s witnesses, Pye, Sislen, and Modne testified that some optical glass is used in the ultraviolet and/or infrared region of the optical spectrum, as is the Schott’s glass involved in this case. Dr. Kreidl, another Schott witness, testified he knew of optical glass that was black.

The court also considered excerpts from Schott’s own catalogues, which described the glass at issue as “colored optical glass.”

In light of the conflicting evidence before the Court of International Trade, we cannot say that the court erred in holding that Schott has failed to show that the decision in Schott I was clearly erroneous.

The judgment of the Court of International Trade is

AFFIRMED.