Precipitation is one of the most important factors of climate, and the analysis of spatio-temporal variability of precipitation is a primary way to understand the formation and development of regional water resources. Furthermore, the trend of the spatio-temporal variability of precipitation directly affects the structure, service function and spatial distribution of various ecosystems. Based on the monthly precipitation data from 1953 to 1987 and from 2007 to 2012, using regression analysis, anomaly, spatial autocorrelation, Mann-Kendall test, Co-kriging interpolation considering the influence of elevation (DEM), and cross-validation, we conducted an analysis on the spatial and temporal variability of precipitation in Dianchi basin. The results are stated in the following sections: (1) from 1953 to 1987, the precipitation tends to increase in spring, autumn and winter, but decrease in summer; from 2007 to 2012, however, the precipitation tends to decrease in all seasons except in autumn. (2) An increasing trend was discovered in the annual precipitation (11.12mm/10a) during the 35 years, and it generally revealed a fluctuation pattern of "down-up-down"; however, a downward trend of precipitation is significant from 2007 to 2012. (3) The indices of Moran's I are negative during the 35 years, which reveals a main negative correlation that is different from the period of 2007 to 2012. The analysis of LISA shows that the spatial heterogeneity tends to change with respect to the geographical location and time. (4) The spatial distribution of annual rainfall is similar to the rainfall during the rainy season, which is a cross-distribution indicated by two high-value and two low-value areas. In addition, the spatial extent of heavy rain events is decreasing, and the spatial extent of light rain events is increasing.