In order to improve the accuracy of microwave remote sensing soil moisture products in China, the data detected by active and passive microwave soil moisture products were fused. Soil moisture data from the China Land Data Assimilation System (CLDAS) was selected as reference to calibrate the deviation. The triple collocation analysis and Pearson correlation analysis were used to calculate the error of microwave remote sensing soil moisture data (ASCAT-A/B, AMSR2, SMOS soil moisture data). And then, weight was calculated according to the error, and the weighted average method was used to fuse the active and passive microwave remote sensing soil moisture data. Compared with ESA-CCI SM, the correlation coefficient is 0.62 and the mean bias error is -0.023 m3/m3. Compared with the ERA-Interim soil moisture reanalysis data, the fused data have high correlations. Compared with the in-situ soil moisture data, the fused data is superior in accuracy with higher correlation coefficient, lower mean bias error and root mean square error to that of ESA-CCI fused soil moisture products and CLDAS surface soil moisture products in China.