AbstractVertical profiles of cloud microphysical properties importantly determine the lifetime and precipitation rate of clouds. The 94‐GHz cloud profiling radar (CPR) onboard the CloudSat satellite can measure the vertical profile of radar reflectivity, from which the microphysical properties of cloud can be retrieved. The retrievals bear variations due to various assumptions and auxiliary products used. This study targets on the mid‐latitude clouds in the northern hemisphere, and intercompares the CloudSat products describing the vertical profiles of cloud microphysics and evaluate the uncertainties for each retrieval algorithm, with further evaluation by aircraft in‐situ observations over the North China Plain region during 2013–2017. For those retrieval products performing phase apportion, the ambient temperature‐based linear apportioning on mixed‐phase clouds can produce reasonable estimation on ice water content, apart from the heavily precipitating clouds. The retrieved liquid water content constrained by cloud optical depth well matched in‐situ observations, however its effective size is overestimated (hereby underestimating the number concentration of water droplets) because of the influence of larger precipitating hydrometeors on size distribution. The CPR‐only retrieval can well produce the effective diameter and number concentration of ice for deep convection clouds, but using additional lidar constraint underestimates the effective diameter due to the intense attenuation by thick clouds. The analysis here suggests the appropriate parameters from various products for different cloud types, and provides guidance for future development of retrieval algorithms on vertical profiles of cloud microphysical properties.