Abstract: Flood disaster is one of the most common natural disasters in China. The research on the risk assessment of regional flood disaster has become a hot topic in the field of disaster science and regional planning. In view of the flood disaster in Jingzhou City,Hubei Province in July 2020,the random forest model is used to complete the measurement of flood impact factor contribution rate and the construction of disaster risk assessment system in Jingzhou City,and then based on XGBoost model,regional flood risk assessment and regionalization are carried out. The results show that:（1）The three indexes contributing most to the risk of flood disaster in Jingzhou City are the buffer zone of river system,flood storage area and night light intensity. Extreme daily rainfall,soil moisture,soil texture,proportion of cultivated land area and rainstorm days in recent 5 years also contributes significantly to flood disaster risk assessment.（2）The performance and prediction accuracy of the risk assessment model reaches the highest after the four indexes of land use type,slope aspect,vegetation coverage and distance from main roads decrease.（3）The area of high risk area of flood disaster in Jingzhou is 15.30%,and the area of high risk area of flood disaster is distributed significantly along Honghu Lake,mainstream of Yangtze River and its subordinate tributaries. Therefore,Jingzhou is very important for the detection of water level of Yangtze River and the planning of the flood dike. Honghu City,eastern Jianli City,Gong’an County and Shishou City are the most vulnerable to the flood disaster in Jingzhou. Relevant management personnel should take timely measures to prevent and reduce disasters.