Two along-track (level 2) satellite precipitation retrievals by the Global Precipitation Measurement (GPM) Microwave Imager (GMI) and the Dual Frequency Precipitation Radar Ku-band (DPR-Ku) and two multisatellite precipitation products, global satellite mapping of precipitation (GSMaP) and Integrated Multisatellite Retrievals for GPM (IMERG), are intercompared for different cloud types during warm season over the western North Pacific region. It is found that the biases of the precipitation measurements are systematically associated with cloud types. The best agreements of passive microwave (PMW) products and infrared-based (IR) products with satellite radar-based estimates are found for a relatively weak precipitation range for mid-low clouds (except over land) and high clouds, while similar agreement is found for heavier precipitation range for deep convection regardless of surface type. Precipitation from mid-low clouds over land is considerably underestimated by PMW and IR products over almost the entire intensity range. The IR-based precipitation estimates for deep convective clouds considerably overestimate the intensity for both weak precipitation and cases where precipitation was not detected by the DPR-Ku algorithm. The findings reveal the characteristics of the biases of the products depend on the associated cloud types, which suggests consideration of the cloud type information to improve satellite-based precipitation estimates.