Global warming and urbanization processes have led to frequent urban waterlogging disasters. Communities, as the basic components of cities, are the fundamental units that bear various disaster incidents. Therefore, it is imperative to improve the level of public safety management and alleviate urban waterlogging disasters in communities. Taking the communities under Wuxing District of Huzhou City as the case study, we constructed an urban waterlogging model at the community scale to simulate and analyze the changing characteristics of waterlogging processes under six different rainfall scenarios. The results show that the variation of surface runoff peak is mainly affected by the land use composition of sub-basins and the average slope. The lower the building area ratio, the higher the road area ratio, and the steeper the average slope is, the faster the response speed of runoff peak becomes. The formation of overflow nodes is affected by pipe diameter and the convergence of rainwater. Under high rainfall return periods, the impact of rainwater convergence is greater. The proportion of overflow nodes increases from 41.48% for a 3-year return period to 56.50% for a 50-year return period, and the node overflow shows different trends under different simulation durations. Buildings and other impervious surfaces bear more surface water accumulation, which mainly distributs in the middle section of Jintian Road and its vicinity, the intersection of Qingtong Road and Longwangshan Road, as well as the surrounding areas of the southern section of Taihu Road.