Given their extensive spatial and temporal coverage, satellite-based remote sensing methods are uniquely suited for global snowfall observations.  However, while satellite data have been extensively used in many cloud and rainfall studies, existing satellite remote sensing techniques do not provide accurate estimates of snowfall over the higher latitudes. In this study, a snowfall retrieval algorithm over land based on Bayes’ theorem is developed using high frequency microwave satellite data. The most important component of the Bayesian algorithm is the a-priori database consisting of high frequency microwave brightness temperatures (> 89 GHz) and corresponding snowfall rates. In this algorithm, observational data from both spaceborne and surface-based radars are used to construct an a priori database of snowfall profiles. For these profiles, combined with NEXRAD radar data, the CloudSat Profiling Radar data are used, which can provide profiles of cloud and precipitation properties. They are then used as input to a forward radiative transfer model to obtain brightness temperatures of microwave satellite sensors. WRF simulations are also used to provide temperature and humidity information for the radiative transfer calculations. The scattering properties of the snowflakes are calculated on the basis of realistic nonspherical shapes using a discrete dipole approximation. Land emissivity is highly variable and complex making it more problematic to use microwave data over land than over ocean. The NOAA Microwave Land Emissivity Model (MEM) is employed to estimate the microwave land emissivity, using the Advanced Microwave Sounding Unit (AMSU) A/B data and AGRMET, which is a land surface model run by the Air Force Weather Agency. The retrieval algorithm is applied to AMSU-B satellite data for snowfall cases occurring over North America. Since the interaction of the microwave radiation with snow-covered surfaces requires another extensive research, here, snowfall cases are carefully selected during early winter seasons or in rarely snow-covered regions to reduce snow-accumulated surface effects. The retrieved snowfall rates using AMSU-B data from snowfall cases are then examined and compared with surface radar observations.