Abstract: Extreme precipitation events have the characteristics of low probability and high risk. The use of high resolution dataset is helpful to understand the real situation of extreme precipitation. Based on the dataset of China's hourly precipitation from 1961 to 2013, the characteristics of the intensity and frequency of extreme precipitation in China were analyzed by various mathematical statistics methods. The results showed that: Firstly, the intensity of hourly extreme precipitation in China at different return periods had obvious spatial differentiation characteristics of high in southeast and low in northwest. The hourly extreme precipitation thresholds in China under different percentiles also had the spatial differentiation characteristics of high in southeast and low in northwest. Secondly, the trend of hourly extreme precipitation frequency in China from 1961 to 2013 under different percentiles were mainly increasing, and with the increase of percentiles, the increasing trend tended to decrease and concentrated in the Yangtze River Basin. The fluctuation of hourly extreme precipitation frequency in China showed a spatial pattern of small fluctuation in the southeast and large fluctuation in the northwest. With the increase of percentiles, the areas with large fluctuation expanded from the northwest to the east and south, while the western areas gradually showed a mosaic pattern of fluctuation small and large. EOF analysis of China's hour extreme precipitation frequency under 90% and 95% percentile showed that the hourly extreme precipitation frequency in China had distinct temporal and spatial variation modes, and the first two modes could reflect the main temporal and spatial variation modes of the hourly extreme precipitation frequency in China. The two methods of super threshold sampling reflected the temporal and spatial characteristics of the Yangtze River Basin.