Abstract: To investigate evolution laws of flood processes and status of flood risk in upstream Huangtaiqiao area of the Xiaoqinghe River basin, Jinan city, a coupled flood model integrating SWMM and InfoWorks ICM-2D was developed． The main parameters of the SWMM model were calibrated using a genetic algorithm with an elite retention strategy． The model was used to simulate and analyze pipeline network drainage capacity, surface inundation status, and hazard distributions under different rainfalls with various return periods． The improved genetic algorithm could quickly optimize model accuracy to a high level in a short time, significantly improving efficiency of parameter calibration． The coupled model exhibited good applicability in the study area． With an increase of rainfall return period, number of overloaded pipes and overflow nodes in the pipeline network increased significantly． Capacity of underground pipeline-network was basically saturated at 10 a return period; the water-logging risk rose rapidly when return period exceeded 10 years． The total inundated area of the river basin increased from 573.38 hm² at a 5 a return period to 956.26 hm² at a 100 a return period, with areas of high water depth concentrated on major roads such as Erhuan Road (S), Jingshi Road, and Lashan Interchange． The flood risk level continuously increased with extension of return period, and area of high-risk areas continued to increase - with return periods from 5 a to 100 a, the area of relatively high-risk zones and high-risk zones increased by 151.11 hm² and 263.50 hm², respectively． This study area showed a certain risk resistance capacity against short-return-period rainfall, but targeted water-logging prevention measures should be prioritized under extreme rainfall conditions．