Topsoil moisture is an important parameter to quantitatively monitor drought, and it plays an important role in the ecological environment in arid areas. The normalized difference vegetation index (NDVI) of threshold methods were used to classify land vegetation types, MODIS data were used to select optical remote sensing algorithm feasible to calculate the benchmark of soil moisture, FY3B/MWRI radiance data and microwave remote sensing algorithm were used to invert daily variation of top soil moisture, and finally the two techniques were combined to setup a cooperative inversion model of soil moisture for northern Tibet. The results showed a significant correlation between the inversed and in situ soil moistures, the determination coefficient was 0.89 and RMSE was 0.97, which indicated the cooperative inversion model was more accurate than the inversion model derived from single remote sensing data, and it is suitable for inversing topsoil moisture in northern Tibet.