Near-bed, highly resolved velocity profiles were measured in the lower 0.03 m of the water column using acoustic Doppler profiling velocimeters in narrow tidal channels in a salt marsh. The bed shear stress was estimated from the velocity profiles using three methods: the log-law, Reynolds stress, and shear stress derived from the turbulent kinetic energy (TKE). Bed shear stresses were largest during ebbing tide, while near-bed velocities were larger during flooding tide. The Reynolds stress and TKE method gave similar results, while the log-law method resulted in smaller bed shear stress values during ebbing tide. Shear stresses and turbulent kinetic energy followed a similar trend with the largest peaks during ebbing tide. The maximum turbulent kinetic energy was on the order of 1 × 10 À 2 m 2 /s 2 . The fluid shear stress during flooding tide was approximately 30% of the fluid shear stress during ebbing tide. The maximum TKE-derived shear stress was 0.7 N/m 2 and 2.7 N/m 2 during flooding and ebbing tide, respectively, and occurred around 0.02 m above the bed. Turbulence dissipation was estimated using the frequency spectrum and structure function methods. Turbulence dissipation estimates from both methods were maximum near the bed (~0.01 m). Both the structure function and the frequency spectrum methods resulted in maximum dissipation estimates on the order of 4 × 10 À 3 m 2 /s 3 . Turbulence production exceeded turbulence dissipation at every phase of the tide, suggesting that advection and vertical diffusion are not negligible. However, turbulence production and dissipation were within a factor of 2 for 77% of the estimates. The turbulence production and dissipation decreased quickly away from the bed, suggesting that measurements higher in the water column cannot be translated directly to turbulence production and dissipation estimates near the bed.