Deterministic data envelopment analysis (DEA) and stochastic production frontier (SPF) models are alternative methods for estimating capacity in fisheries. Fishery managers should be aware of likely differences in the capacity estimates obtained from these approaches if such estimates are to be used to support capacity reduction programs. In this paper, we provide a comparative analysis of DEA and SPF capacity estimates for a variety of possible capacity concepts using a panel data set for 10 vessels in the U.S. Northwest Atlantic scallop fishery. We find that DEA capacity output measures are higher than corresponding SPF measures, but that the two approaches provide similar guidance about overall and even relative boat-specific capacity levels under certain circumstances. The variations that emerge suggest, in particular, that biases can arise from inferring capacity output at "efficient" production levels, which disregards customary and usual operating conditions.