A suspended-sediment-induced, stably stratified oceanic bottom boundary layer is examined with the Mellor-Yamada level II turbulence closure model. The boundary layer equations are coupled through the eddy diffusion coefficients for momentum and mass. The bottom boundary layer of the Florida Current is modeled using flow and sediment properties as input parameters. Model results indicate that the boundary layer response to a suspended sediment concentration gradient is to reduce appreciably the level of turbulence. Turbulent kinetic energy and eddy conductivity are reduced by about 40% and 25%, respectively, and bottom stress is reduced by approximately 45%. Increases of veering angle of ?5¿, together with enhancement of the Ekman-like nature of the current spiral with sediment-induced stability, are noted. The boundary layer adjusts to the reduced turbulence level by a decrease in the slope of the velocity profile and by diminution of the thicknesses of the boundary layer and contained logarithmic layer by as much as 50%. The predicitions are consistent with findings from studies of stably stratified atmospheric boundary layers, and sediment-carrying channel flows. The velocity field near the bottom of a neutrally stratified bottom boundary layer is governed by the relationship U = (u*k)1nz/z0. To recover the predicted velocity distribution from this relationship in the sediment-stratified case, von Karman's constant &kgr; must be reduced by about 15%. The model, however, treats &kgr; as an internal constant. Similarity theory suggests that a more appropriate expression for the near-bottom velocity field in the sediment-stratified bottom boundary layer is U = (u*/&kgr;)(ln z/z0+AFzozdz/L(z)), where the Monin-Obukhov length L = z/Rf,A = 5.5,and Rf is the flux Richardson number. For the thermally stratifed atmospheric boundary layer, L(z) is nearly constant, which yields the familiar log-linear profile. Our model predicts that in the sediment-stratified bottom boundary layer, L is nearly linear in z (i.e., Rf?const). This yields U?(u*k')lnz/z0, where &kgr;' = &kgr;/(1+ARf) and &kgr;' may be interpreted as a modified von Karman constant. The results suggest that u* values inferred from logartihmic speed profiles in sediment-laden flows may be significant overestimates of the actual u*.