<p>Urbanization and climate change have significant impacts on urban hydrological processes, leading to frequent urban flooding, which seriously affects urban development and residents’ life. Therefore, timely and effective potential flood hazard assessment is essential for disaster warning. However, previous assessment methods were limited by more complicated processes and data requirements, resulting in poor regional applicability. To efficiently and reasonably realize the potential flood hazard assessment and further analyze the distribution characteristics of flooding in different cities, this study proposes an assessment approach by integrating runoff yield and confluence mechanisms. Specifically, the Soil Conservation Service Curve Number (SCS-CN) model is used to calculate surface runoff yield based on impervious surface percentage and precipitation data. Cellular automata (CA) are then used to simulate urban runoff confluence through overland flow rules. Finally, the proposed method is applied to potential hazard assessment and distribution analysis in eight Chinese cities, such as Beijing and Zhengzhou. The results indicate that the method is suitable for multi-scenario potential flood hazard assessment in different regions, showing high accuracy and rationality. In addition, there are differences in hazard level distribution across different regions. Among them, cities with large built-up areas and flat terrain show a continuous distribution pattern of potential hazards. In contrast, cities with diverse patterns and complex terrain demonstrate a scattered distribution pattern. Therefore, the results can guide urban flood warnings, as well as the design and construction of sponge city.</p>