This paper analyses the static convergence and dynamics of regional income distribution in rural China . The convergence of income distribution during 1989—2007 is examined by kernel density estimation and Silverman's multimodality test . It is found that there's no statistically significant'twin peaks'convergence in the regional income distribution . However ,there exists a dynamic change in it . The results of factor analysis have the following two indications . First , between 1993 and1998 ,the change of wageincome and household businessincome resultedin the dynamic process of income distribution from'single peak'to'twin peaks', and back to'single peak'a gain .Especially , the change of household business income contributed to a'twin peaks'income distribution in the earlier stage and a'single peak'one in the later stage while the change of wage income always led toa'single peak'd istribution .Second ,thechange of wageincome had been the main driving force for maintaining'singe peak'distribution since 1999 . The impact of wage income on the income distribution could be decomposed into a distributional effect and a growth effect ,in which the latter ,namely the growth of average regional income ,was the key to maintaining the‘single peak’income distribution in rural China .