This paper reports on a numerical model developed to simulate vertically and horizontally polarized microwave emission from the Earth in the frequency range between 5 and 50 GHz and at various angles of observation, with a 1¿¿1¿ (latitude by longitude) spatial resolution, taking into account seasonal variations. The principal motivation of the model is the evaluation of the noise antenna temperature of telecommunications satellites, which is required to calculate the uplink G/T for satellite-borne receivers. The results of the study, however, prove useful in a number of remote sensing applications. To implement the model, significant types of surface, such as bare soil, nonarboreous vegetation, forests, snow, glacier and sea ice, and ocean, have been identified, and their emissivity properties have been determined by the available theoretical and/or empirical models. The millimeter-wave propagation model of Liebe <1993> has been used to compute the atmospheric contribution. Profiles from actual radio soundings collected during a 10-year period over the globe have been used to take into account major climatic variations. The various contributions from the surface and the atmosphere have been finally combined to obtain the theoretical global brightness temperature of each 1¿¿1¿ pixel. The numerical model has been validated by comparing on a pixel-by-pixel basis the theoretical brightness temperature with those measured by the special sensor microwave imager (SSM/I) radiometer in the year 1992 at 19.35, 22.235, and 37.0 GHz at the available polarizations. The discrepancies between model and experimental brightness temperatures have been noted, and actions have been taken to reduce the differences. In its present configuration, the global emission model yields brightness temperature estimates which differ all over the Earth by less than 12 K rms from those measured by the SSM/I. ¿ 1998 American Geophysical Union