Knowledge of migration rates of bedforms provides an indirect indication of the behavior of tidally averaged bottom currents, enables optimization of hydrographic survey frequency and may enable calculation of bedload transport rate. To measure bedform migration rate, we test the use of spatial correlation as a measurement method, which quantifies and locates a region of maximum similarity between two spatial variables. For the latter, we use consecutive eight-bit images of spatial gradient, derived from bathymetric digital terrain models, carrying out the correlation over this representation of the shape of the seabed rather than the bathymetric surface. The digital terrain models were compiled from six repeat multibeam surveys of a headland-associated bank near Saint John, New Brunswick, with a roughly 30-day interval. Vectors are drawn depicting the movement of a sand dune at time t0 toward a point in the spatial correlation array at a later time, t1. A number of different techniques of picking the end of the migration vector were used. The sinuosity of the dune crest at the scale of the correlation window has an impact on which migration vector is the better pick. Averaging of migration vectors from consecutive epochs diminishes random errors in the correlation picks using any single pair of images and creates a more accurate picture of the migration field. Migration rates and crest-relative migration directions vary substantially across the sand bank, reflecting the high gradients in bottom shear stress around the headland.