High-resolution stratospheric and mesospheric temperature measurements from the Cryogenic Infrared Spectrometers and Telescopes for the Atmosphere (CRISTA) experiment taken during the space shuttle mission STS 66 in November 1994 show large tidal signatures in the day/night temperature differences. Previous comparisons with the predictions of the global-scale wave model (GSWM-95) for equinox conditions have shown these differences to be in very good qualitative agreement with GSWM results for the diurnal component of the migrating solar tide, although quantitative differences in the tidal amplitudes and phases did exist. In this paper the source of these differences is examined. Background fields corresponding to the flight conditions are derived from the CRISTA data and used as input to an updated version of the GSWM. The updated GSWM includes revised tidal heating and dissipation schemes. The background fields from CRISTA include temperature, ozone, pressure, mass density, and derived geostrophic wind (in geostrophic balance with the CRISTA temperatures). These model updates significantly improve the agreement with the CRISTA observations at all latitudes and altitudes. Modeled and observed phases match to within 1 hour at the equator. The amplitudes agree within 1 K for most altitudes. Above 75 km, however, the CRISTA measurements show a decrease in the tidal amplitude which is not present in the GSWM results, suggesting that the model underestimates the equatorial gravity wave dissipation in the form of eddy diffusivity and wave stress in the upper mesosphere. The mesospheric wind field update accounts for the most profound improvement in the phase distributions of the diurnal tide. The results show the strengths of the GSWM but also indicate the need for some model improvements. Our determination of the GSWM capability for reliable predictions of phases and amplitudes of the migrating diurnal tide when realistic atmospheric background conditions are used, provides some guidance in the way such models should be used in the future. In addition, the understanding of the diurnal cycle of chemical active species requires an accurate tidal prediction, due to the temperature dependencies of the reactions and the transport processes associated with the tides. ¿ 2000 American Geophysical Union