The topographic accuracy of the AN/APQ-102(XA-2) side-looking radar and its specific applicability to 1:50,000 and 1:250,000 scale topographic mapping were tested using stereo radar techniques with real data. The two basic stereo configurations, opposite-side and same-side, were tested. Also tested was the geometric fit of each individual radar record to the established ground control. Although original plans called for a test area 25 miles long by 4 miles wide, shortcomings in the data actually acquired led to adoption of a much smaller test area. Moreover lack of certain collateral data required development of a special reduction technique based upon ground control in the test area. Photogrammetric aerotriangulation of aerial photography flown simultaneously with the radar provided a means of defining the flight path. Ground points in the test area were also aerotriangulated, and used as control and as the standard for measuring radar accuracy. The test results were surprisingly good, with the opposite-side stereo configuration providing the better accuracy. Average root mean square (RMS) values of 7.7 meters, 12.1 meters and 13.2 meters in X(across-track), Y(along-track) and Z(elevation), respectively, were obtained for the opposite-side case; corresponding RMS values of 9.5 meters, 20.0 meters and 16.7 meters were obtained for the same-side case. Results from fitting the individual radar records to the aerotriangulated positions also demonstrated excellent geometric fidelity in the radar imagery.