To investigate the influence of rainfall and tide superposition on the process of urban waterlogging in coastal cities, the GPU accelerated surface water flow and transport (GAST) model coupled with the storm water management model (SWMM) was used to simulate the characteristics of urban surface water accumulation and the process of pipe network drainage in the central urban area of Haishu District, Ningbo City, under different combination scenarios. The results indicate that the scenario with the largest peak waterlogging area occurred under a 50-year return period, high tidal level, and rainfall peak preceding the tidal peak, where the peak waterlogging area reached 119 800 km², 1.2 times the minimum area. The peak waterlogging area under overlapping rainfall and tidal peaks was consistently smaller than under staggered peaks. Under the same return period, the peak flow rates at typical outlets Ⅰ, Ⅱ, and Ⅲ were significantly higher during high tidal levels compared to typical tidal levels, and the peak flow rates at all three outlets increased with higher return periods, with their peak occurrence times advancing gradually. The overlapping of rainfall and tidal peaks significantly reduced the peak flow rates at the outlets, with the greatest reduction observed under the 50-year return period scenario.