A current profiler consisting of a vertical array of three electromagnetic current meters has been evaluated through an intercomparison of the three sensors, with reference to nearby current and wave data and by comparison to recent laboratory performance tests (Aubrey and Trowbridge, 1985). Mean flow estimates are too uncertain and variable to allow bottom boundary layer shear stress to be estimated by the conventional logarithmic law method. As well as unexplained sudden shifts in the mean speed response, the comparison with vector-averaged current meter data indicates possible long-term reduction in response due to fouling of the sensors by biological growth. The directional response was less sensitive to fouling effects. The oscillatory response on one occasion after field deployment for 17 days indicates a reduction in response from 41 to 45% at a period of oscillation of 3 s in a combined steady and oscillatory flow field. This study demonstrates that despite careful laboratory calibration, electromagnetic current meters are not at present suitable for quantitative study of dynamics of sediment resuspension in near-bottom shallow-water environments. ¿ American Geophysical Union 1987