A comparison is made between real aperture radar (RAR) measurements and simulations (based on modeled tidal currents) of radar cross section over a complicated tidal basin (in the vicinity of the Phelps Bank region of the Nantucket shoals) in order to more fully understand the origin of radar signatures that are observed at the ocean surface as a consequence of variations in the topography of the ocean bottom. The Phelps Bank region was mapped under two extreme wind speed conditions: in high winds, in excess of 15 m/s, and in low winds, of the order of 2--3 m/s. For the light-wind case the measured radar cross section over the west side of the Phelps Bank was enhanced by as much as 20 dB relative to the clutter background. For the high-wind case, no discernible bathymetric signature was found in the high-clutter background. Numerical results for the two-dimensional M2 (semidiurnal) tidal currents over the Phelps Bank (Greenberg et al., 1989), with 1/8¿1/8 min of arc resolution, are used as input to the surface signature models: the Alpers and Hennings (1984) first-order Bragg relaxation model; a generalized form of this relaxation model (in which wind directional effects are incorporated in an approximate manner); and the full-spectrum model of Lyzenga and Bennett (1988). Comparisons between the models (which do not include wave breaking) and an extreme case of 2--3 m/s winds (where strong wave breaking could become important) reveal that although the models predict correlation between variations in bottom topography and surface signature, they significantly underpredict the magnitude of the observed effect. The model calculations also are very sensitive at low (<2 m/s) wind speeds to the functional form that is assumed for the wind-wave forcing in the wave action equation. Prior visual observations and measurements of wave spectra (and wave shoaling) in the vicinity of Phelps Bank strongly suggest that the deficiencies of the modeled results that occur explicitly at light winds are due to wave breaking. A number of additional experiments and measurements are suggested for more normal environmental conditions for further theory assessments. ¿ American Geophysical Union 1994