Fish predation has been recognized to influence community composition and structure, but has not been represented in experimental community tests of chemical effects. To investigate the feasibility of including vertebrate predation in 65-L microcosms, four species of small fish (Gasterosteus aculeatus, Cottus asper, Poecilia reticulata, and Pimephales promelas) were tested. These studies suggest that predation can be included in moderately sized microcosms only if exposure is restricted either by limiting the time of fish presence or by allowing the fish access to a limited portion of the microcosm. Predation was simulated by filtration of fixed proportions of the microcosms; no major changes were observed, which suggests that microcosm communities can tolerate removal of at least 25% of selected zooplankton populations per week. Fish predation did not alter the ability to detect direct effects of a selective algicide (streptomycin) on the algae, but did reduce the ability to detect indirect effects on cladoceran and ostracod populations.