The study explored the impacts of rainfall and land use on surface runoff, taking the area surrounding the North Moat in Beijing, China, as an example. We selected 15 rainfall events occurring from April to November in 2011 and 2012, representing six rainfall levels, which were light rain, moderate rain, heavy rain, rainstorm, heavy rainstorm, and torrential rainstorm, respectively. Surface runoff depths during each rainfall event in the seven control sections were simulated using the calibrated SWMM (Storm Water Management Model). The simulated results showed that the modeled runoff depth significantly linearly increased with increasing rainfall. There were distinct variations in runoff depth when rainfall varied between different rainfall levels. Although rainfall and runoff depth might exhibit different dynamics over time during different rainfall events, runoff depth at each time was determined by rainfall in the previous 1-2 h, and the overall variation in runoff depth depended on the early cumulative rainfall. Runoff depth significantly declined with increasing permeability area ratio (PARO), occurring a critical threshold of 15%-20% at control section scale. PARO had greater impacts on the runoff in rainstorm, heavy rainstorm, and torrential rainstorm events. The result can provide important theoretical basis and implications for catchment management and urban planning to reduce surface runoff and flood peak and avoid waterlogging disasters, which are very common in current China.