Abstract：This paper analyzes the historical rainfall data of Shaoxing National Meteorological Observatory, by using methods of linear regression, wavelet analysis and Mann-Kendall mutation testing, to reveal the seasonal variation and interannual change of 48 year's rainfall series in Shaoxing region., and to forecast the future rainfall. The results are as follows. Multiple time scale features can be seen in the whole seasonal and annual rainfall series. The 3~4 years period variations prevail in spring rainfall series, with a transition trend from 3 years period to 4 years period. 4 years and 14 years period variations are the main types in summer rainfall series. 2 years and 12 years period variations are mainly found in autumn rainfall series, and 4 years and 16 years period variations in winter rainfall sequence. 2 years, 4 years and 13 years period variations prevail in annual rainfall sequence. The sudden change of spring, winter and annual rainfall is not obvious. The mutation points of summer and autumn rainfall occur, respectively, in 1973 and 1963. The same tendency is seen in summer and annual rainfall sequence. The summer rainfall mutation influences the annual rainfall mutation. The rainfall of Shaoxing region will become affluent in the future 5~6 years according to the main period of summer and annual rainfall series.