Seed mussel stock assessments rely heavily on techniques similar to those used in seabed mapping and can involve both acoustic data collection and physical ground truthing. Optimum allocation analysis has recently shown value in using acoustic data variances as surrogates of seabed heterogeneity to allocate efficiently the ground-truthing effort that minimizes survey coefficients of variance. Using acoustic data from a single-beam ground discrimination system, optimum allocation analysis was used to direct ground truthing between seed mussel strata. The resulting stock assessment values and variances were compared with actual fished quantities and were found to be highly comparable when compared with random and expert judgment scenarios.