Abstract: Based on the Chinese meteorological stations precipitation data and high resolution surface precipitation data in data set CRU, we used three indexes, including variation coefficient, persistence of above/below average conditions and abrupt change, to study the variation of monthly precipitation in China from 1982 to 2000. Three indicators represent variation in quantity, time structure, and jump change. The results showed that the monthly precipitation varied largely, the variation amplitude in winter was greater than that in summer. The monthly precipitation has no long-term trend and the variation was very significant. The precipitation abrupt changes mainly occurred in July and September, and the least in November. In 1983, the phenomenon of El Nino leaded to the most significant year of precipitation abrupt change. The results of two precipitation data sets are slightly different in different variation indexes. The station data is clear in the variation of precipitation, and CRU data can reflect the actual situation of the whole country in the phenomenon of abrupt changes.