Quantachrome Corporation appeals the May 23, 2000 decision of the U.S. District Court for the Southern District of Florida holding after a bench trial that U.S. Patent No. 5,074,146 (“the T46 patent”), owned by Micromeritics Instrument Corporation, was not proven invalid for obviousness in light of the prior art. Quantachrome also appeals the district court’s summary judgment ruling that the T46 patent is not unenforceable due to inequitable conduct. Because we conclude that U.S. Patent No. 4,083,228 to Turner (the “Turner patent”) and the Autopycnometer 1320 are pertinent prior art references that should not have been excluded from the district court’s consideration during the obvious*849ness determination, the portion of the court’s judgment that the ’146 patent is not invalid for obviousness is vacated. Because the record does not reflect evidence of intent by Micromeritics to deceive the U.S. Patent and Trademark Office (“PTO”), the district court’s summary judgment of no inequitable conduct is affirmed. We remand for a redetermination of obviousness.
Discussion
In its March 29, 2000 summary judgment order, the district court ruled that the Turner patent and the Autopycnometer 1320 (and its manual) are not relevant prior art. The district court restated this conclusion in its May 23, 2000 order following a bench trial on obviousness. Accordingly, the court did not consider these references in determining whether the claimed invention was proven invalid for obviousness.
Micromeritics disclosed both of these references to the PTO as prior art during the original prosecution of the ’146 patent, and they are listed as prior art on the cover of the ’146 patent. Micromeritics also listed both these references as prior art in its December 26, 1996 request for reexamination.
The asserted claim of the T46 patent, claim 27, claims a static volume pycnometer. Elsewhere the patent specification states that “[cjurrent gas comparison pycnometers suffer from many deficiencies which limit the precision of the volume determination. First, the devices are not stable against ambient temperature variations.” ’146 patent, col. 1, II. 38-41. Thus, the 146 patent recognizes that temperature variations may adversely affect volume measurements in pycnometers.
“If a reference disclosure relates to the same problem as that addressed by the claimed invention, ‘that fact supports use of that reference in an obviousness rejection.’ ” In re GPAC Inc., 57 F.3d 1573, 1578, 35 USPQ2d 1116, 1120 (Fed.Cir.1995) (quoting In re Clay, 966 F.2d 656, 659, 23 USPQ2d 1058, 1061 (Fed.Cir.1992)). The Turner patent teaches a solution to the problem of maintaining thermal consistency between the two chambers of a pycnometer by disclosing that the chambers should be formed in a block “preferably made of aluminum or other suitable material of high thermal conductivity to obtain rapid temperature equilibrium between the chambers.” Turner patent, col. 7, II. 11-19. Similarly, the manual for the Autopycnometer 1320 teaches that the sample chamber and the reservoir chamber “are machined within a single aluminum block to ensure temperature uniformity.” These two references relate to the same problem addressed by the claimed invention—i.e., maintaining thermal stability in a pycnometer. It is true that these two references relate to dual-piston pycnometers, whereas the ’146 patent claims a static volume pycnometer. However, the largely unrefuted testimony of Quanta-chrome’s expert, Dr. Athanassios Panagiotopoulos, indicated that a person of ordinary skill would have looked to dual piston pycnometers in finding ways to improve thermal stability in static volume pycnometers. In light of this testimony, the similarity of the problems addressed in all the disclosures', and the fact that Micromeritics disclosed these references to the PTO as prior art relevant to the claimed invention, we hold that the subject matter taught by the Turner patent and the Autopycnometer 1320 manual is sufficiently analogous to that of the claimed invention that the trial court should have considered *850these references during the bench trial on obviousness.
We make no determination as to whether the claimed invention is invalid for obviousness in light of the Turner patent and the Autopycnometer 1320 (and its manual). We merely direct the trial court to reconsider its nonobviousness determination in light of these references, alone, together, or in combination with other prior art.
As to the other issues raised by Quanta-chrome on appeal, we find no grounds for disturbing the trial court’s determinations. As to the Gulf Oil and Quantachrome pycnometers, substantial evidence supports the trial court’s conclusion that these devices were kept confidential, and thus were not prior art. While it is certainly true, as Quantachrome points out, that references that do not qualify as “prior art” under 35 U.S.C. § 102 or § 103 may nonetheless be relevant to obviousness as a secondary consideration, nothing in the opinion below suggests that the trial court held otherwise.
Concerning inequitable conduct, we find no basis in the cited statements in the Camp and Orr declarations to upset the trial court’s summary judgment concluding that Quantachrome failed to show by clear and convincing evidence that Micromeritics intended to deceive the PTO.
Conclusion
Because we conclude that the Turner patent and the Autopycnometer 1320 should have been considered by the trial court during the bench trial on obviousness, we vacate the portion of the trial court’s judgment that the asserted claim of the T46 patent is not invalid for obviousness, and we remand for a redetermination of obviousness. The trial court’s judgment that Quantachrome failed to prove inequitable conduct is affirmed.