Octopus birnaculatus, a common member of intertidal and subtidal communities in southern California, consumes many motile benthic invertebrate species, including snails, chitons, h p e t s , bivalves, and crustaceans. An analysis of octopus d d holes in empty snail shells indicates that 0. bimaculatus is a major s n d predator in the community. The addition of marked snails to a natural reef with unmanipulated octopus densities demonstrated that octopuses can quickly reduce s n d abundances, consuming 24 % of the available snails within 24 d. During 5 yr at 1 site, octopus abundance declined from 40 to 8 octopuses, while total prey abundance increased from 16 to 80 ind m-'. A second site experienced a similar drop in octopus density and increase in prey density over 3 yr. At both sites, snails and hermit crabs showed the greatest increases. Sedentary grazers, bivalves, and crabs and shrimps occurred at low densities throughout, possibly because the abundances of these preferred prey were depressed by octopus predation even during years of low octopus density. In splte of the strong negative association at the 2 sltes between octopus and prey over time, for a single time penod at 15 different s~t e s prey densities were not negatively correlated w t h the densities of their predators. However, octopus densities were much higher when they appeared to reduce the abundances of their prey than when the 15 different sites were sampled. I conclude that octopus predation can dramatically reduce prey densities when octopuses are unusually common, but that in most years the abundance of octopuses 1s too low to disrupt the basic patterns of prey abundance that result from other processes. The hghest prey species richness and diversity at the 2 study sites occurred during years of lowest octopus densities. Species richness, chversity and evenness of the prey assemblage were not correlated with octopus, lobster or predatory gastropod abundances at the 15 different locations; when the effects of habitat structure were held constant by partial correlation, seastar abundance was negatively correlated with species nchness and diversity. The relation between predation and species chversity suggests that, when predation is effective in this assemblage, it decreases diversity.