Quantitative analyses of remote sensing data acquired with variable geometries and different sensors are greatly facilitated by geocoding the data to a common geographical reference. Errors in geocoded SAR images are due to inaccuracies in the orbit data, errors in the processing parameters, and DEM errors. Automated terrain geocoding of SAR data is possible using only a single resampling of the native slant-range to the final map geometry after measurement and correction for residual geolocation errors. An accurate adaptive resampling algorithm for the interpolation from the SAR geometry to the map coordinate system is presented. This processing paradigm has been extended to permit combining data from different SAR instruments.