The vegetation opacity parameter is a key input needed to map surface soil moisture and other landsurface properties to brightness temperature. An integrated approach to estimating vegetation and soil moisture may provide a better soil moisture estimate than relying on opacity estimates from visible/near infrared vegetation indices. We test a new online vegetation parameter estimation method using AMSR-E satellite observations over Little Washita watershed during several growing seasons. An ensemble Kalman filter consisting of a multi-layer soil hydrology model and an omega-tau type radiative transfer model is used to assimilate C-band and X-band, vertical and horizontal polarization observations. A state augmentation approach which appends the parameter(s) to the state vector is used for parameter estimation. Open loop filter statistics (without observation updates) and filter soil moisture statistics using AMSR-E data are computed using an in situ soil moisture dataset available starting from July 2002 in Little Washita watershed. Results show that vegetation parameter estimates are reasonable. Observation bias correction parameter estimation tests within the ensemble Kalman filter are also discussed. This research is relevant for applications directly assimilating brightness temperature observations such as the planned NASA SMAP (Soil Moisture Active Passive) Level 4 surface and root zone soil moisture product.