Based on the cloud products from Japanese Himawari-8 satellite, this paper analyses the precipitation data form 08:00 BT to 17:00 BT in summer (June-August) of 2017 over central and eastern China. The relationships between cloud optical depth (COD), cloud-top particle size (CPS), cloud-top temperature (CTT) and precipitation were discussed in detail. Experiments showed that the rainfall probability is highly correlated to cloud parameters. The rainfall probability increases as the COD and CPS increase, and/or CTT decreases. However, the correlation between single cloud parameters and precipita-tion intensity is not high. The correlation coefficient between COD and hourly rain rate is 0.2315, that between CPS and hourly rain rate is 0.1823, and that between CTT and hourly rain rate is -0.2235, which are all weak correlations. If two or three cloud parameters are combined to form an hourly rain rate distribution matrix, the precipitation process can be more clearly reflected. The case on 28 August 2017 indicates that, compared to the pure infrared-based algorithm, the tri-parameter method can significantly improve the estimation accuracy of hourly rain rate.