Based on hourly precipitation data from 109 meteorological stations in Sichuan basin and 3-hourly forecast data with 0.1°×0.1° spatial resolution from GRAPESMESO model from May to September during 2017-2018, this study investigated the characteristic of key physical parameters of the occurrence and development of extreme shorttime heavy precipitation (hourly precipitation amount ≥50 mm), such as thermal instability, water vapor and dynamic conditions. A prediction model was established using random probability thought and principal component analysis method to develop the extreme shorttime heavy precipitation probability forecast product. The forecast effect evaluation showed that the TS was 24.0% when the probability was greater than or equal to 0.7. The value could be used as the reference threshold for the extreme shorttime heavy precipitation forecast. During the rainstorm process in Sichuan basin on 22 July 2019, the practical application indicated that the product had good reference meaning for the forecast of extreme shorttime heavy precipitation location.