Based on DEM digital elevation model, RUSLE model combined with GIS, GeoDa, GS+ and other software, the evolution of spatial pattern of soil erosion and its influencing factors in 25 years in Chongli District, Zhangjiakou City, Hebei Province were analyzed. The results showed that:(1) From 1990 to 2015, the soil erosion was severe in the central, western, and southwestern regions of the study area. The intensity of soil erosion was mainly mild and moderate. The amount of soil erosion firstly decreased and then increased. (2) From 1990 to 2000, the change of soil erosion intensity was dominated by mild transformation. Soil erosion was alleviated. From 2000 to 2010, it was mainly slight erosion turning to high-level erosion, and showed aggravate trend. From 2010 to 2015, the overall transformation was in the form of slight erosion, mild erosion turned to high erosion levels, but the area of increased erosion decreased, and the erosion slightly improved. (3) Soil erosion's Moran's I>0, the spatial distribution was positive correlation, showing aggregation state, with high-level aggregation mainly in central and western and southwest. Soil erosion modulus met index model and spherical model, R² was 0.943~0.979. The range of soil erosion reduced from 3 870 m to 860 m and then increased to 1 470 m, indicating that the change of soil erosion was firstly fast and then slow down from 1990 to 2015, the spatial distribution of spatial correlation expanded and the spatial heterogeneity increased firstly and then decreased. The fractal dimension of soil erosion ranged from 1.922 to 1.971. Spatial heterogeneity of soil erosion was mainly caused by stochastic factors such as vegetation cover, land use types, and water and soil conservation measures in a small spatial scale. (4) The contribution rate of the first, second, and third principal components to the soil erosion factors accounted for 89.215 0%. In the first principal component load, the projection length of vegetation cover factor was the largest, which was 0.976 4. In the second and third principal component loads, the projection length of soil and water conservation measures factor and soil erodibility factor were large. Therefore, the influencing factors of soil erosion in Chongli District were in order of vegetation cover factor, soil and water conservation measures factor, soil erodibility factor, rainfall erosivity factor, and slope length slope factor. The results can provide theoretical basis for soil and water management and sustainable development of Qingshui River Basin in Chongli District.