This study was conducted to determine the effects of otter trawl tow length on variahility and catch rates, and to assess replicate sampling. Longer tows caught more individuals/minute, individuals/tow, and species/tow than the shorter tows. The total number of species caught per hour of sampling, however, were essentially equal for all tow lengths. No difference in precision hetween tow lengths were detected. The numher of replicates needed to detect changes within desirahle statistical limits were logistically impractical. We recommend that otter trawls he used strictly for qualitative purposes and conclude that a small number of replicated lO-minute tows would adequately descrihe hasic community structure.