Abstract: Using the daily precipitation data of 16 national meteorological stations in Southeastern Guizhou from 1961 to 2020, the small-medium rain, heavy rain, rainstorm and heavy rainstorm in daytime, nighttime and daytime during flood season (April to September) were selected as the research objects. And the temporal and spatial distribution characteristics of rainy days at all levels were compared and analyzed by using EOF decomposition, polynomial fitting, linear trend and other methods. The results show that the rainy days at all levels in the flood season of Southeastern Guizhou was characterized by uneven distribution characteristics, the heavy rainfall at night was mainly concentrated in the west, and the heavy rainfall in the daytime was concentrated in the east. Under the control of large-scale weather system, the number of rainy days at all levels at night and during the day showed a consistent phase change. The anomalies of small to moderate rain at night in Taijiang, heavy rain at night in Kaili, heavy rain at night in Sansui, small to moderate rain at day in Huangping, heavy rain at day in Tianzhu, and heavy rain at day in Danzhai were high, and the regional characteristics were significant. The interannual changes of rainy days at all levels of day and night were similar, and the number of days above heavy rain had increased significantly after entering the 21st century. The number of rainy days at all levels at night in each month was more than that in the daytime. The heavy precipitation at night and daytime was concentrated in June and July, and the number of rainstorm days at night was over 50% more than that in the daytime. The number of light to moderate rain days at night and during the day decreased significantly, and the number of days above heavy rain increased linearly. At some stations, the number of heavy rain or heavy rain at night increased significantly. The number of rainy days at all levels at night showed a significant short-period change, and the number of rainy days at all levels during the day included 1~2 significant short, medium and long-period changes.