We experimentally reduced densities of predatory fish in replicated 2 m2 areas of the littoral zone in two ponds to test whether density and biomass of invertebrates would respond to release from fish predation. The ponds are of similar size and in close proximity, but support different fish assemblages: bluegills (Lepomis macrochirus Rafinesque) and largemouth bass (Micropterus salmoides (Lacepede)) in one pond, and bluespotted sunfish (Enneacanthus gloriosus (Holbrook)) and chain pickerel (Esox niger Lesueur) in the other. Fish densities were reduced to less than 15% of ambient levels in both experiments. In the bluegill–bass pond, density and biomass of most invertebrate taxa and size classes were unaffected by the fish manipulation. Total invertebrate densities did not differ significantly between fish treatments, but total invertebrate biomass was significantly greater where fish density was reduced, averaging 30% higher over the course of the study. Likewise, manipulation of fish in the bluespotted sunfish–pickerel pond had few significant effects on individual taxa and size classes. There were no significant effects on total invertebrate abundance in the bluespotted sunfish–pickerel pond. Our results provide direct experimental evidence consistent with the collective evidence from previous work, suggesting that the impact of fish predation on density and biomass of invertebrate prey in littoral habitats is variable, but generally weak. Invertebrates that coexist successfully with fish in littoral systems probably are adept at taking advantage of refugia offered by the structurally complex physical environment.