Abstract: In September 2019, many control stations in Yi-Shu-Si River Basin have experienced the largest flood since 1974, which is affected by Super Typhoon Lekima and cold air in 2019. Aiming at the rain flood process, this paper uses many methods, such as real-time rolling forecast, historical similarity flood comparison, peak and storage correlation and multi-system forecast, to forecast flood accurately of the main control stations and lakes in the watershed, which provides decision-making basis for flood control and dispatch in Yi-Shu-Si River Basin in time.