Rainfall erosivity is one of the key parameters for soil erosion, sediment yield and water quality modeling, and has grown in importance for environmental modeling of climate change. Based on daily precipitation data from the 5 meteorological stations the annual rainfall erosivity was calculated with the 8 models and the difference between the models were analyzed. The results that the maximum deviation is Model H and its error is 76.2%,the minimum deviation is Model C and its error is 2.1%. The relative amplitude of Model C is closest to the average, which indicates Model C is more suitable for the study area than other models. Monthly rainfall erosivity models are superior to the daily and annual models in the Bailongjiang River Basin. Annual mean erosive rainfall of annual precipitation in Diebu, Dangchang, Zhouqu, Wudu and Wenxian is 46.1%,44.8%,43.6%,46.0% and 34.5%, respectively. And the relationship between annual rainfall and annual mean rainfall erosivity is highly correlated, fitting power function. Simultaneously, the correlation coefficient of each meteorological station is more than 0.98.