Prior to the study we report here it had been thought that the sole undisturbed nighttime mechanisms for producing CO2 vibrationally excited in the &ngr;3 mode were (1) vibrational excitation by thermal collisions and (2) absorption of 4.3-&mgr;m earthshine by CO2. In this paper we show detailed evidence that the mechanism OH(&ugr;)+N2→OH(&ugr;-1)+N2≠, followed by N2≠+CO2→N2+CO2(&ngr;3), may be the dominant excitation mechanism for producing CO2(&ngr;3) in the 85-km altitude region. The evidence is based on 4.3-&mgr;m zenith radiance data obtained on April 11, 1974, via a rocket-borne liquid N2 nitrogen cooled circular variable-filtered radiometer. The rocket was launched at night from the Poker Flat Research Range, Alaska. The data were obtained under conditions of essentially zero auroral activity. There is a feature in these data near 85-km altitude which can be explained by the mechanism OH(&ugr;)&ugr;&ugr;N2≠&ugr;&ugr; CO2(&ngr;3) →CO2+h&ngr;4.3&mgr;m. A column transfer of 0.12¿0.025 erg/cm2 s from OH≠ to N2 is required to explain the feature. An alternate explanation on the basis of just the mechanisms 1 and 2 which are cited above requires that there existed a very unlikely mesospheric temperature profile on the evening of April 11, 1974. A similar nonauroral feature appears consistently near 85 km in preliminary 4.3-&mgr;m zenith data obtained on March 27, 1973, February 25, 1974, March 6, 1975, and March 12, 1975, thus providing further evidence that the feature most likely results from the OH mechanism rather than a strange mesopheric temperature profile. These data were obtained from the Air Force Geophysics Laboratory and Utah State University as part of the Infrared Chemistry Experiments, Coordinated Auroral Program supported by the Defense Nuclear Agency. In this article we confine our attention to analysis of the April 11, 1974, data.