Comparison of photochemical calculations of atmospheric ozone between 40 and 55 km with measurements from a satellite-borne remote sensor shows acceptable agreement in the high-latitude summer hemisphere. However, in the sunlit high-latitude winter, the available data imply either a smaller water vapor mixing ratio than generally accepted or a temperature 15¿--20¿K colder than contained in published model atmospheres. High-latitude measurements of the 1.27-μ column emission rate of O2(1Δg) carry information on lower-mesospheric ozone and are well reproduced by theoretical calculations for summer conditions. As with the ozone data the infrared emission in winter implies an odd oxygen loss rate smaller than predicted on the basis of standard water vapor and temperature models. The magnitude of the 1.27-μ signals and their consistency with upper stratospheric ozone data cast serious doubt on the large mesospheric ozone abundances deduced in independent experiments.