An ensemble Kalman Filter-based data assimilation framework that links a crop growth model with active and passive (AP) microwave models was developed to improve estimates of soil moisture (SM) and vegetation biomass over a growing season of soybean. Complementarities in AP observations were incorporated to the framework, where the active observations were used to optimize surface roughness and update vegetation biomass, while passive observations were used to update SM. The framework was implemented in a rain-fed agricultural region of the southern La-Plata Basin during the 2011-2012 growing season, through a synthetic experiment and AP observations from the Aquarius mission. The synthetic experiment was conducted at a temporal resolution of 3 and 7 days to match the current AP missions. The assimilated estimates of SM in the root zone and dry biomass were improved compared to those from the cases without assimilation, during both 3-and 7-day assimilation scenarios. Particularly, the 3-day assimilation provided the best estimates of SM in the near surface and dry biomass with reductions in RMSEs of 41% and 42%, respectively. The absolute differences of assimilated LAI from Aquarius were <0.29 compared to the MODIS LAI indicating that the performance of assimilation was similar to the MODIS product at a regional scale. This study demonstrates the potential of assimilation using AP observations at high temporal resolution such as those from soil moisture active passive (SMAP) for improved estimates of SM and vegetation parameters.