Abstract: The significant uncertainties in cloud parameterization in multiple global climate models (GCMs) need to be addressed based on better understanding of temporal and special variation of clouds. Global and zonally averaged vertical distributions of cloud occurrence frequency of eight cloud types in three phases and their seasonal variations obtained from the combined CloudSat radar and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) lidar measurements from March 2007 to February 2010 are investigated. The global distribution pattern of cirrus is found to be similar to that of deep convective clouds, which are mainly distributed over the warm pool of the West Pacific, the monsoon region and the intertropical convergence zone (ITCZ). The seasonal pattern of cirrus and deep convective clouds vary with pressure and wind zone. Stratocumulus and stratus are largely located over non-monsoon regions in the middle and low latitudes and over the oceans in high latitudes. Altocumulus and altostratus show obvious difference between land and sea. Cumulonimbus and cumulus exhibit distinct differences between the high and low latitudes. The spatial pattern of ice phase clouds is similar to that of cirrus while the pattern of water phase clouds is similar to that of stratus and stratocumulus. Zonally averaged vertical distribution of clouds indicates that the average cloud top of ice phase clouds decreases with increasing latitude. Water phase clouds are located in low altitudes (about 2 km). Mixed phase clouds are mainly found over an arcuate zone ranging from 0-10 km. Stratiform clouds have a higher probability to display a multiple-layer structure, while convective clouds are prone to form as single-layer cloud clusters. Cloud overlapping is more obvious for stratiform clouds than for convective clouds. Distributions of cumuliform and stratiform clouds are consistent with distributions of convective and stratiform precipitation. These results are valuable for evaluation of cloud diagnostics in global and regional climate models.