Soil erosion is one of the most serious soil degradation problems in the world. Its abnormally occurrence, development and change process directly affect the development and utilization of water and soil resources. Meanwhile, it threatens human survival, social and economic development. Numerous studies have shown that the increased vegetation coverage was a main factor to reduce soil erosion. It was an important issue attracted much attention of researchers in soil and water conservation that the quantitative relationship between vegetation coverage and soil erosion in different research areas in the world. Ten Turbutries located in northern Ordos is one of the most serious soil erosion areas, which was mainly distributed by moderate and severe erosion area, and covered by sandstone loess. In our study, the soil erosion modulus was calculated using Universal Soil Loss Equation (USLE) based on remote sensing technique, rainfall, DEM and land use/cover data. Then, the relationship between vegetation coverage and soil erosion modulus was analyzed by the method of overall fitting and sectional fitting regression to identify the vegetation coverage corresponding to the threshold of soil erosion modulus. The results showed that the annual average soil erosion modulus was 29.31 t hm^-2a^-1 from 2000 to 2017. The largest in 2016 and the smallest in 2011 were 65.6 t hm^-2a^-1 and 10.95 t hm^-2a^-1, respectively. The change trend of soil erosion modulus was a typical parabolic curve (P<0.001). When the slope grades were <5°, 5-10°, and > 10°, the threshold of soil erosion modulus were18.18 t hm^-2a^-1, 34.29 t hm^-2a^-1 and 74.56 t hm^-2a^-1, and the corresponding vegetation coverage were 11.42%, 16.51% and 16.5%, respectively. In the sandstone loess area, the soil erosion modulus not only would be affected by the factors in USLE model, but also limited by the amount of soil erodibility. In addition, it should be paid great attention to the years with large amount of erosion when judging the relationship between soil erosion and vegetation coverage in regional scale.