The precipitation and surface conditions in the semi-arid areas of northern China have significant spatial and temporal heterogeneity. How to completely and accurately describe the hydrological process in such areas is one of the challenges in contemporary hydrological research. Spatiotemporal precipitation schemes were constructed with different time steps and numbers of rainfall stations in Suide and Caoping Watersheds. To provide a reference for the development of hydrological models in semi-arid regions, large-scale numerical simulation experiments were conducted to explore the influence of precipitation spatiotemporal heterogeneity and runoff generation mechanisms on flood simulation. The results show that:1) The runoff generation in semi-arid areas is strongly influenced by the rainfall intensity in small and medium-sized watersheds, so the time step of precipitation has a great influence on the flood simulation accuracy. In contrast, the impact of the increase or decrease in the number of rainfall stations on rainfall is mainly the absence of storm recentres and small differences in the average rainfall, which has a small effect on the flood simulation. 2) The ability of the hydrological model to accurately describe the dominant hydrological processes is the key to good flood simulations in semi-arid areas. The scale and spatial heterogeneity of the watershed is a priority for model development in such areas. Regardless of the combination of time step and rainfall station density, the model simulations with suitable runoff generation structures always tend to be better.