REDD (Reduced Emissions from Deforestation and forest Degradation) constitutes a set of financial incentives designed to reduce CO2 emissions from forest degradation and deforestation. REDD success depends on measuring forest biomass as a proxy for CO2 stocks. We tested the efficacy of airborne X- and P-band interferometry as a remote-sensing method to quantify forest biomass and detect changes in forest structure in the Paragominas region, eastern Amazon. With field-classified regions of interest (ROI) and radar imagery, we classified an area of 1479.66 km2 into four forest classes. Radar backscatter and interferometric variables of each forest class were statistically examined. We obtained the interferometric height, Hint, by subtracting digital elevation models resulting from X and P band interferometry for the study area. Inventory-measured biomass were obtained for 42 plots nested within these forest classes, and used as ground truth for subsequent analyses. Using these field plots as experimental units, a functional relationship between radar variables and above-ground biomass (AGB) was obtained by fitting a linear model relating inventory-measured AGB and radar-derived variables. A map of AGB was created. Combining backscatter variables and Hint effectively classified the forest. AGB can be predicted for the Paragominas landscape with Hint and the P-band polarizations pHV and pVV (R2=0.82, normalized RMSE=13.7%). X- and P-band airborne radars can be used to estimate forest AGB of large continuous areas as well as detecting forest exploration and degradation, events REDD intends to prevent.