Abstract: To assess the risk of rainstorm disasters in the densely populated areas of eastern Qinghai and enhance the ability to prevent rainstorm disasters. Based on the daily precipitation data of 25 national meteorological stations in eastern Qinghai from 1961 to 2021, combined with geographic information, remote sensing and socioeconomic data, methods such as the percentile threshold method, index system risk assessment method, and entropy weight method were used to analyze the characteristics of rainstorm weather in eastern Qinghai and construct a rainstorm disaster risk assessment model. Finally, the actual disaster conditions of rainstorm processes in the recent 3 years were used for evaluation and verification. The study found that: (1) From 1961 to 2021, the number of rainstorm processes, daily maximum precipitation, cumulative precipitation in the process, and process duration in eastern Qinghai all showed an increasing trend; (2) The intensity index of rainstorm processes in eastern Qinghai shows an overall increasing trend, with two high-value areas in the northeast and south.; (3) The overall distribution of rainstorm disaster risks in eastern Qinghai is characterized by high risks in the northeast and low risks in the west. Among them, the Xining urban area, Datong County, Huangyuan County, Huangzhong District, Menyuan County, Huzhu County, Ledu District and Minhe County are high-risk areas for rainstorm disasters; (4) The test in the recent 3 years showed that 72.04% of the townships with disaster conditions appeared in high and higher-risk areas, and only 13.17% appeared in lower and low-risk areas of rainstorm disaster risks. The rainstorm disaster risk assessment model well reflected the disaster situation caused by rainstorm weather in eastern Qinghai..