Techniques were developed to extract radar scattering coefficient (&sgr;0) signatures associated with the internal wave patterns observed in the SAR Internal Wave Signature Experiment (SARSEX) data set. The SARSEX experiment was conducted in the New York Bight region of the Atlantic Ocean in August-September 1984. A theoretical model was exercised which uses environmental conditions and radar imaging parameters to calculate the changes in the synthetic aperture radar (SAR) image intensity due to internal waves. For conditions present during SARSEX, this model predicts that the modulations should be higher at L band than at X band, the modulations should decrease at both frequencies as incidence angle increases, range-traveling waves should have higher modulations than azimuth-traveling waves, and azimuth-traveling internal waves should have broader signatures than range-traveling waves. The analysis of the SAR-observed internal wave patterns showed that the X band internal wave modulations were comparable in magnitude to those found at L band, the range-traveling waves had higher modulations than azimuth-traveling waves, and the expected decrease in modulation with increasing incidence angle was present at times for the range-traveling cases but not present in the azimuth-traveling cases. Finally, comparisons showed that the SAR-observed modulations were still somewhat underpredicted by the theoretical model, especially for the azimuth-traveling internal wave case. ¿ American Geophysical Union 1988