We examine the direct effects of errors in determining temperature, pressure, and density in the background atmosphere on the comparison of ozone measurements by two different satellite sensors. Such comparisons, particularly between satellite instruments with different fundamental measurement techniques, provide an important method for checking the performance of these sensors. For examples, ozone abundance profiles from the solar backscattered ultraviolet (SBUV) spectrometer on the Nimbus 7 satellite are being reanalyzed to remove the effects of long-term instrument drift. Measurements from other ozone instruments are vital to this reanalysis: a primary comparison is that between SBUV observations and those from the Stratospheric Aerosol and Gas Experiment II (SAGE II) instrument on the Earth Radiation Budget satellite (ERBS). SAGE II observes ozone by solar occultation, giving values of ozone concentration versus geometric altitude. SBUV provides ozone overburden and its derivatives versus pressure. The conversion between geometric altitude and pressure requires accurate information on the vertical structure of the background atmosphere. In this study we compare atmospheric structure measurements routinely available from the National Meteorological Center with a special set of in situ measurements by radiosondes and meteorological rocketsondes. Furthermore, we demonstrate the manner in which the differences in these background atmosphere measurements propagate through the direct comparisons of stratospheric ozone profiles by SBUV and SAGE II sensors. Modest differences in vertical positioning can propagate into significant ozone differences in regions with strong vertical ozone gradients, particularly at 70 mbar in the tropics. We find apparent ozone differences of 5--10% or more due solely to these background atmosphere effects within an analysis for which the ozone observations themselves are treated as error-free. ¿ American Geophysical Union 1991