Temperature measurements in the stratosphere and mesosphere were taken during the first Cryogenic Infrared Spectrometers and Telescopes for the Atmosphere (CRISTA 1) mission using CO2 emissions. These measurements range from 13 to close to 100 km, and individual temperature measurements have a precision of 1 K. The CRISTA orbit was circular at an inclination of 57¿, so local time variations during the 7 day mission were small for a given latitude and orbit leg. Zonal averages of the data show significant structure in the vertical and as a function of latitude. Temperature differences between the zonal mean data from the ascending and descending portions of the orbit are of the form expected from the diurnal tide. The maximum zonal mean difference is ≈20 K and occurs over the equator at an altitude of 75 km. Zonal variations in the temperature difference indicate that the tidal amplitude is not uniform at all longitudes. At the equator the maximum amplitude (30 K) appears over the African sector and the minimum (10 K) over the Pacific sector. This variation is most likely a nonmigrating wavenumber 2 diurnal tide although other less plausible possibilities exist. To facilitate comparisons with model results the temperature variations are converted to equivalent vertical displacements, assuming the tidal motions are adiabatic. Such an approach is appropriate in the mesosphere where diurnal variations due to radiative effects are small and reduces the dependence of the measured temperature amplitude on the background temperature profile. Equivalent vertical displacements are also calculated using results from the global scale wave model. For the most part the model results and observations are in excellent qualitative and good quantitative agreement from the tropopause to the mesopause. However, the observed vertical wavelength is smaller, and its amplitude in terms of equivalent displacement is smaller in the upper mesosphere than that in the model. The decrease in amplitude is consistent with that expected from the transition from equinox to solstice conditions. ¿ 1999 American Geophysical Union