Suspended sand concentration profiles and current speed measurements were made in 1.6 ± 0.4m water depth on the seaward side of a bar beyond the break point at Stanhope Lane beach, Prince Edward Island, Canada, during a 4 day period. Two mild storm events occurred and 7 min-long bursts of 4.4 Hz data were recorded approximately every hour. Concentration profiles, with a resolution of 5 mm and 7.5 mm, were obtained using a 3 MHz acoustic backscatter (ABS) instrument. Some data on bed form dimensions were obtained by divers between storm events. Multiple bed echoes from the ABS also provide further information on the ripple heights but no direct observations were available at times of higher waves. During the 4 day period the local position of the bed decreased by about 8 cm, erosion occurring in two short (⋟6 h) periods, one during each storm event. Mean currents at this location were weak (<0.1 m s −1). The resuspension coefficient γ 0, calculated from the concentration at 2 cm above the bed, decreased as the wave height increased (in the break-off range) supporting the observations of Vincent et al. [(1991) Marine Geology, 96, 1–18]. Large inter-burst variability was observed in the concentration profiles and in the eddy diffusivity and suspended transport fluxes computed from these profiles. This variability was due to the short length of record (⋟7min) relative to wave groups and to the location of the ABS relative to bed forms; to obtain consistent concentration and transport fluxes it is necessary to average many bursts over a time scale that is long compared to both groupiness and bed form mobility. Except at the beginning and end of events when wave conditions were changing rapidly, averaging over groups of bursts with similar wave conditions produced eddy diffusivity profiles characterized by a linear ɛ s gradient of (20 ±2.5) cm s −1 from the sea bed to 20 ± 5cm, beyond which ɛ s slowly decreased, and a suspended sand transport which was all below 20 cm and dominated by a jet-like shoreward flow in the lowest 3–4 cm above the sea bed. Bursts obtained when conditions were in, or close to, the equilibrium range of bed forms showed concentration profiles which were partly exponential (hence ɛ s was constant with height) and transport profiles with more structure. Sensitivity tests indicate that concentrations and eddy diffusivities may be underestimated by about 20% by using the modal size for the sand in suspension rather than a distribution of sizes. Towards the end of the second storm a “low suspension event” of ⋟6 h duration occurred and suspended sand concentrations about 5–10 cm decreased by 1–2 orders of magnitude while the near bed concentrations remained approximately constant, suggesting that the bed forms responsible for vortex ejection of sand higher into the water column had been wiped out and the sea bed had become (temporarily) flat. Wave heights did not change significantly but wave period increased from 4.8 to 6.2 s, decreasing the maximum bed shear stress. The low suspension period may have a transient response of the sea bed to changed hydrodynamic forcing prior to the subsequent growth of bed forms with a different wavelength.