Synthetic aperture radar (SAR) imagery showing two types of features near the inshore edge of the Gulf Stream are compared with nearly simultaneous estimates of the sea surface slope field derived from optical shipboard measurements. One class of feature consists of a set of narrow, dark lines having radar signal modulations of about -10 dB at L band. These modulations are comparable to those observed in the in situ wave-slope data over wavenumbers corresponding to the radar-resonant Bragg waves. These modulations are also predicted by a simplified equilibrium wave spectrum model using a surface elasticity of about 25 mN/m (as determined from in situ measurements of surface tension) and a wind friction velocity of about 10 cm/s (from buoy and shipboard measurements). These results support Lyzenga and Marmorino's <1998> inference that the Gulf Stream dark lines represent relatively passive, biogenically derived surfactant slicks that are advected and strained by the large-scale surface flow field. The second class of feature consists of a pair of bright-dark signatures over the shallow continental shelf region near Cape Hatteras, North Carolina. These features are characterized by signal modulations of about +4 and -4 dB in both the L band SAR and in situ data. These modulations appear to be due to the interaction of surface waves with the current convergence and divergence regions associated with a rotary circulation in a shallow convergent front. Model calculations of this interaction by Lyzenga <1998> indicate that the positive perturbations associated with current convergence should be relatively independent of the wavenumber, which is in approximate agreement with the in situ measurements of the present paper. The model also indicates that the negative perturbations associated with the region of current divergence should decrease for wavenumbers higher than L band. That trend is observed in the SAR data but not in the in situ data, which show significant perturbations at higher wavenumbers. It is conjectured that this is due to the effects of surfactants and either along-front or time variability in the frontal dynamics, which are not included in the model. Additional comparisons of this kind should be useful to further test and improve physical models of near-surface phenomena. ¿ 1999 American Geophysical Union