Abstract: Runoff is the main carrier of soil erosion. Accurate prediction of runoff is crucial for soil erosion forecast and reservoir safety management in the Three Gorges Reservoir. To optimize the SCS-CN model’s calculation method for soil antecedent moisture content and assess the model’s applicability before and after improvement in the Three Gorges Reservoir area, this study introduced the antecedent precipitation index to characterize soil moisture and set an upper limit threshold based on measured soil saturated moisture content, thereby improving the SCS-CN model. Using the rainfall-runoff data from the Zhongxian Soil and Water Conservation Monitoring Station from 2019 to 2022, the applicability of the model before and after improvement under different rainfall types and land uses was evaluated. The results show that among the two initial models and two improved models, the improved model based on MSCS-CN（M4） has higher simulation accuracy for runoff depth, and the Nash efficiency coefficient（NSE） is improved by 134% compared with the original SCS-CN model（M1）, especially for the case of saturated soil water content in the antecedent period where the NSE is improved to 0.74, and the introduction of the maximum antecedent precipitationindex（APImax） as the upper limit of the antecedent precipitation index can reduce the overestimation of the predicted values of high runoff events in previous models. In addition, the improved model M4 is well applicable to the main flow-producing rainfall types（high rainfall intensity） and the main land uses（croplands and orchards） in the Three Gorges Reservoir area, and effectively improves the runoff simulation accuracy for the small rainfall-producing events and the land use types with high erosion risk（the Nash efficiency coefficients for Rain Type II, croplands, and orchards are 0.66, 0.69, and 0.75, respectively）. Overall, the improved model M4 improves the accuracy of runoff simulation, and is more suitable for the runoff generation mechanism and land use types of the Three Gorges Reservoir area, which can provide a valuable reference for runoff prediction and soil and water loss prevention and control work in the Three Gorges Reservoir area.