Accuracies of the benthic acoustic stress sensor (BASS) and the proposed modular acoustic velocity sensor (MAVS) are evaluated. A simple model of the hydrodynamic sources of error for acoustic current meters is presented and compared with the measured performance of BASS in a tow-tank and in field deployments. The sources of error addressed include those due to ideal flow around the sensor cage, wake effects from the support structure and transducers, vortex shedding from the cage, and electronic zero-point offsets. Electronic error dominates at velocities less than 5-10 cm/s, while flow disturbance dominates at higher speeds. >