The SMAPVEX12 (Soil Moisture Active/Passive Validation Experiment 2012) was carried out over the summer of 2012 in Manitoba, Canada. The goal of the project was to improve the accuracy of satellite-based remote sensing of soil moisture. Data were gathered during a 42-day field campaign with surface measurements on 55 different agricultural fields in south-central Manitoba. The extended duration of the campaign, contrast in soil textures, and variety of crop types over the study region provided an excellent range of soil moisture and vegetation conditions. The study fields ranged from bare to fully vegetated, with volumetric soil moisture levels spanning a range of almost 50%. Remotely sensed data were collected on 17 days by aircraft at 1.4 Ghz with a microwave radiometer at two different resolutions. Observed brightness temperatures from the radiometer showed a typical inverse relationship to the near simultaneous soil moisture measurements from the field. Field-by-field relationships using all sampling dates with both soil and emissivity data were all shown to be significant with the exception of two of the pasture fields and a soybean field. Linear regressions across multiple fields and by flight lines also had statistically significant slopes. The significance of all these relationships improved with the removal of pasture fields from the analysis. On most fields, the sensitivity (slope) of the relationship and correlation coefficient (R2) between emissivity and observed soil moisture increased when vegetation and roughness effects were taken into account. The b parameter that relates vegetation water content to optical depth in the tau-omega model was optimized using the collective slope and R2 values of the individual fields. A b parameter value of 0.06 for horizontal polarization and 0.13 for vertical polarization were found to be optimal across the range of all fields in this analysis.