The Soil Moisture Active Passive (SMAP) satellite provides global soil moisture products with reliable accuracy since 2015. However, significant gaps of SMAP soil moisture appeared over Tibetan Plateau. To address this issue, we proposed two methods, machine learning and geostatistics technique to fill the spatial gaps of SMAP L3 soil moisture. For the machine learning technique, we train a Random Forest algorithm which aims to match the output of available SMAP L3 soil moisture using a series of input variables such as SMAP brightness temperature (TBH, TBV) in ascending orbits, surface temperature, MODIS NDVI, land cover, DEM and other auxiliary data. Then, the established RF estimators were applied to the SMAP brightness temperature from descending orbits to reconstruct complete soil moisture data over the Tibetan Plateau. For the geostatistics technique, the Ordinary Kriging was applied to the available SMAP L3 soil moisture pixels to interpolate complete soil moisture data. To cross-validate the performances of the algorithms, we assume certain areas with available SMAP SM values as missing, and then compared the gap-filling results with the actual ones. The cross-validations show that the gap-filling results from two algorithms were highly correlated to the SMAP official SM products with high coefficients of determination (R2_RF = 0.97, R2_OK = 0.85) and low RMSE (RMSE_RF = 0.015 cm3/cm3, RMSE_OK = 0.036 cm3/cm3). Furthermore, the gap-filling soil moisture data present a better correlation with the SMOS soil moisture data (R = 0.55 ~ 0.7) than the GLDAS simulations (R = 0.18 ~ 0.62).