Abstract: By using 3S technology, and based on the analysis and prediction of the land use change in Baimahe basin, a SWAT model was established to study the runoff response of the basin under the scenarios of different land use. In the meantime, the contribution coefficients of the main land use types in the basin to the runoff depth were calculated. From 1987 to 2017, the main land use types in the basin were farmland, construction land, forestland, and shrub land, occupying 96% or more of the total land area, while the grassland, waters, and unused land only had a smaller proportion. The four main land use types had different effects on the runoff depth. The contribution coefficient of forestland, shrub land and construction land to the runoff depth was 2.61, 0.38, and 0.34 mm·km^-2, respectively, implying that these three land use types had positive effects on the runoff depth in this basin. On the contrary, the contribution coefficient of farmland was -0.11 mm·km^-2, implying that farmland had negative effect on the runoff depth.