Abstract: To investigate the spatio-temporal variation patterns of extreme precipitation in the Poyang Lake Basin, ten extreme precipitation indices were calculated based on daily observational data from meteorological stations within the basin from 1961 to 2022. A comprehensive evaluation function for extreme precipitation was constructed using principal component analysis, which was then used to score and rank each station. The spatial distribution characteristics were further elucidated by K-means cluster analysis. The results indicated that: ① Most stations showed an increasing trend in extreme precipitation indices, except for some in the south showing a decreasing trend. The historical averages of all indices in the northern part were generally higher than those in the south, except for the Consecutive Dry Days (CDD)index. ② The various extreme precipitation indices were highly correlated. Three principal components were extracted through principal component analysis, with the cumulative variance contribution rate accounting for 96% of the total variance, effectively reflecting the main characteristics of extreme precipitation. ③ All stations were classified into three clusters via cluster analysis. The low-value cluster was primarily distributed in the southern part of the basin, while the high-value cluster was concentrated in the northeast, and the classification was highly consistent with the actual observations.