In order to compare and evaluate the differences and causes of different passive sensor ice cloud data in China, and to find out the products that are closer to the actual characteristics of ice clouds, this paper selected three sensor data include ice cloud variables: MODIS (Moderate-resolution Imaging Spectroradiometer), AIRS (Atmospheric Infrared Sounder), AIRS/AMSU-A (Atmospheric Infrared Sounder/Advanced Microwave Sounding Unit-A) from JAN 2013 to DEC 2015 and DARDAR (raDAR/liDAR) active sensor data was used to verify. The results show that ice clouds occurrence frequency overall horizontal distribution of MODIS in China is closer to DARDAR data, but the numerical analysis of AIRS and AIRS/AMSU-A in different small-scale areas in southern, southwest, northeast and central China is better than MODIS; The optical optical thickness AIRS/AMSU-A is similar to the DARDAR data inversion, and the overall national variation law is also the AIRS/AMSU-A data in China. The AIRS passive sensor that large-scale migration changes of ice cloud effective particle radius in China and the surrounding is similar to DARDAR data. In numerical studies, MODIS should be selected for small areas at mid-high latitudes, and AIRS data should be closer to DARDAR at lower latitudes; MODIS passive sensor can more accurately describe the three years’ seasonal changes regardless of ice cloud occurrence probability, optical thickness or ice particle radius.