Abstract: In order to understand the spatial distribution characteristics of rainfall erosivity in Yimeng Mountain Area, the authors used the ordinary kriging and regression kriging to predict the rainfall erosivity spatial distribution and to analyze the prediction precision and differences of the different interpolation methods, based on the 88 rainfall stations’ daily precipitation data (1980—2010) in Yimeng Mountain Area. The results indicated that, 1) although the spatial distribution regularities of rainfall erosivity in this two methods were coincident, the regression kriging was better for describing the local variabilities. The rainfall erosivity interpolation results of regression kriging are close to the actual calculation values, compared with the ordinary kriging, and the relative prediction accuracy was improved by 53.64%. 2) The maximum value of rainfall erosivity in Yimeng Mountain Area was 5 438.22 MJ·mm/ (hm²·h·a) and the minimum value was 3 033.23 MJ·mm/ (hm²·h·a); the mean value was 4 133.92 MJ·mm/ (hm²·h·a). The analysis of semi variance function showed that the average annual rainfall of flood season and the annual erosive rainfall were the main factors affecting the spatial variability of rainfall erosovity. 3) The spatial distribution pattern of rainfall erosivity in Yimeng Mountain Area decreased gradually from the south to the northwest and from the south to the northeast. The spatial distribution of rainfall erosivity in Yimeng Mountain Area was medium variability, but it was weak variability at county scale (except Zoucheng).