Abstract: Based on the annual rainfall series in the Taihu Lake basin and seven sub-areas during the period of 1951—2011， three heterogeneity indices such as the apportionment entropy disorder index， Gini coefficient and variational coefficient have been applied to analyze the heterogeneity of the annual and rainy season rainfall distribution at the monthly and decadal scales. The relationships between the historical flood or drought and the heterogeneity indices are investigated. The long term changing trend and constancy of rainfall in the homogeneity indices are tested by the trend-free pre-whitening Mann-Kendall and Hurst. Also， its possible influences on future drought or flood have been primarily discussed in this study. Analysis results show that there are small differences among the three kinds of the rainfall heterogeneity indices. When the annual rainfall and heterogeneity index values are both high， it is easy for the study area to suffer flood disaster. When the annual rainfall is relatively low while the index values is high， it is easy to suffer drought. With the heterogeneity of rainfall distribution in the rainy season increases significantly， the risk of drought and disasters is also rising in the coming years. So it is necessary to continue to analyze precipitation heterogeneity in the space in order to reveal precipitation differences in the space scales. The study would provide a technical support to disaster prevention and disaster reduction for the basin.