Accurately predicting soil loss is crucial for the ecological environment protection in drylands. Biological soil crusts (biocrusts), which are widespread surface coverings in dryland ecosystems, significantly reduce soil loss from slopes. However, the existing soil erosion models often inadequately account for the effects of biocrusts, which may lead to overestimation of soil loss in areas covered by biocrusts. Accordingly, we conducted 18 simulated rainfall experiments with a rainfall intensity of 90 mmh(-1), using 10 m x 2.1 m plots with biocrusts of different coverage on a 25 degrees slope to explore the relationship between the soil loss ratio (SLR) and biocrust coverage. Subsequently, we incorporated the biocrust factor (B-BSC) into the biological-control (B) factor of the Chinese Soil Loss Equation (CSLE). Meanwhile, we used the soil loss data from six-year natural rainfall to investigate the relationship between soil loss and biocrust coverage under natural rainfall conditions and verified the revised CSLE. The results showed that soil loss decreased with the increase in biocrust coverage under natural rainfall conditions, and the degree of decrement was related to the rainfall amount. There was a significant negative exponential function relationship between the SLR and biocrust coverage. Based on this, the B-BSC was developed as a sub-factor of the B factor, with the formula B-BSC = 0.8417e(-0.0688BC), BC was biocrust coverage (%) (R-2 = 0.880, P < 0.01). Compared with the original CSLE, the prediction accuracy of the revised model was significantly enhanced. This study verified the importance and practicality of revising the erosion model so as to consider the effects of biocrusts on soil loss at plot-scale and provided valuable reference for the revision of erosion models in drylands covered by biocrusts, such as the Loess Plateau in China.