Microwave observations from past and present satellites (such as Scanning Multichannel Microwave Radiometer-SMMR, Tropical Rainfall Monitoring Mission Microwave Imager - TMI, Advanced Microwave Scanning Radiometer on NASA EOS Satellite Aqua - AMSR-E, Microwave Radiometric Imager on Fengyun-3B satellite of China Meteorological Administration - MWRI, the WindSat of Naval Research Lab, and SMOS of European Space Agency) have been used to retrieve land surface soil moisture data products for many years. However, different retrieval algorithms using the same tau-omega radiation transfer equation and the same satellite observations have produced very different soil moisture data products. One of the main causes is the handling of the masking of microwave signals from soil by the spatially heterogeneous vegetation cover. Vegetation optical depth (VOD) is directly related to the dielectric properties and water content of the vegetation and crucial for soil moisture retrieval accuracy. In this paper, two parameterizations of the VOD are examined for the single channel retrieval algorithm: one is based on a microwave vegetation index (MVI) and the other is based on the Normalized Vegetation Difference Index (NDVI). Using surface soil moisture data from the North-America Land Data Assimilation System (NLDAS) and the AMSR-E observations, the VOD for each 0.25 degree grid of NLDAS domain can be inversed with the tau-omega radiation transfer equation. The MVI and NDVI can be obtained from AMSR-E and MODIS and their empirical relationships to the inversed VOD can be established using the regression approach for different types of vegetation cover, respectively. For soil moisture retrieval operations, these empirical VOD parameterization equations are used to estimate VOD from either MVI or NDVI. Then the SCR algorithm is used to retrieve soil moisture from the X-band observations of AMSR-E or MWRI. The VOD retrievals can be evaluated against the inversed VOD values for both MVI and NDVI approaches while the soil moisture retrievals from the SCR algorithm are compared with the original NLDAS estimates. The relative performances of the MVI and NDVI approaches for the VOD parameterization will be presented. The soil moisture retrievals from AMSR-E and MWRI using the SCR algorithm are compared with in situ soil moisture measurements. These soil moisture retrieval will also be compared with the soil moisture retrievals from the Land Parameter Retrieval Model.