Abstract: In the context of global climate change and rapid urbanization, extreme rainstorm events have become more frequent, and urban waterlogging issues are increasingly severe, threatening urban safety.To mitigate the risks of waterlogging and improve emergency management during extreme rainstorm events, analyzing urban waterlogging processes under such conditions using simulation methods has become an important research trend.Based on an analysis of the basic characteristics of extreme rainstorms, the key factors influencing urban waterlogging were identified.A systematic review on the two main simulation methods for urban waterlogging under extreme rainstorm conditions was conducted: mechanism-driven models and data-driven models.The former explicitly represent physical processes but require longer computation times, while the latter meet the demands of fast simulation and forecasting but lack physical mechanisms.Based on this, future development trends for urban waterlogging simulations under extreme rainstorm conditions were discussed from five perspectives: multi-index dynamic analysis of simulation results, improving simulation accuracy and efficiency, deep integration of city-scale and watershed-scale models, dynamic integration of mechanism-driven models with numerical weather forecasting, and real-time coupling of mechanism-driven and data-driven models.The research findings can provide valuable references for identifying and managing urban waterlogging processes under extreme rainstorm conditions.