Striped bass Morone saxatilis and hybrid bass M. saxatilis x M. chrysops have been stocked to establish fisheries in many US reservoirs, but success has been limited by a poor understanding of relations between prey biomass and predator growth and survival. To define sizes of prey that are morphologically available, we developed predictive relationships between predator length, mouth dimensions, and expected maximum prey size; predictions were then validated using published data on sizes of clupeid prey (Dorosoma spp.) in five US reservoirs. Further, we compared the biomass of prey considered available to predators using two forms of a length-based consumption model - a previously published AP/P ratio and a revised model based on our results. Predictions of maximum prey size using predator GW were consistent with observed prey sizes in US reservoirs. Length of consumed Dorosoma was significantly, but weakly, correlated with predator length in four of the five reservoirs (r2 = 0.006-0.336, P < 0.05). Model predictions of available prey biomass differed by as much as 800% between the original AP/P model and a revision based on our estimates of maximum available prey size. The revised model predicted less available prey biomass in cases where large Dorosoma (>150 mm TL) were abundant. (C) 2000 Elsevier Science B.V.