Using one year of CloudSat radar data, the characteristics of the vertical profiles of snow water content/snowfall rate have been examined. Whether or not a radar reflectivity profile corresponds to surface snowfall is determined by ECMWF surface temperature and an empirical formula (relating precipitation phase and surface air temperature) derived from surface station observations. The radar reflectivity (Z) is converted to snowfall rate (S) and ice water content (IWC) by Z-S/Z-IWC relations derived with consideration of the nonsphericity of the ice particles. The backscattering coefficients of the nonspherical ice particles are simulated by discrete dipole approximation. The vertical distributions of snowfall rates/ice water contents are analyzed by grouping the data according to surface snowfall intensity, surface type, weather condition and latitudinal regions. In the presentation, I will first explain the retrieval method, and then discuss the 2-D (height and snow intensity) PDFs of the snowfall rate/ice water content, their variation with regions, surface type (land/ocean), and synoptic conditions. This study is important not only for understanding the global snowfall climatology, but also for developing a snow cloud database used for retrieving snowfall from satellite high-frequency passive microwave observations.