An algorithm for retrieval of spatial variations of the mean square slope (MSS) of the sea surface from sunglint imagery is proposed. The retrieval algorithm is free on a prior suggestion on PDF model. The transfer function, relating the brightness contrast to the MSS contrasts, is found from observed sunglint brightness, where “real” PDF of the sea slopes has built-in. Developed approach was applied for analysis of the sunglint signature of the mesoscale ocean dynamics and internal waves. We found that the ocean currents (eddies, meanders, frontal lines) and internal waves are well visible on the sea surface as the MSS anomalies. Results of this study is further adopted by Kudryavtsev et al. (2010) for development of advanced approach for synergetic use of SAR and optical imagery in studies of meso-scale ocean dynamics.