The Soil Moisture and Ocean Salinity (SMOS) satellite has opened the era of soil moisture products from passive L-Band observations. In this study validation of SMOS products over continental US is done by using the SCAN/SNOTEL soil moisture monitoring stations. The SMOS operational product and the SMOS reprocessing product are both used and compared over year 2010. First, a direct node to site comparison is performed by taking advantage of the oversampling of the SMOS product grid. The comparison is performed over several adjacent nodes to site, and several representative couples of site-node are identified. The impact of forest fraction is shown through the analysis of different cases across the US. Also the impact of water fraction is shown through two examples in Florida and in Utah close to Great Salt Lake. A radiometric aggregation approach based on the antenna footprint and spatial description is used. A global comparison of the SCAN/SNOTEL versus SMOS is made. Statistics show an under estimation of the soil moisture from SMOS compared to the SCAN/SNOTEL local measurements. The results suggest that SMOS meets the mission requirement of 0.04 m3/m3 over specific nominal cases, but differences are observed over many sites and need to be addressed.