Mesocosm tests provide important data for evaluating ecological risk in the aquatic environment. In 1988, the Office of Pesticide Programs provided guidance concerning the conduct of such tests. Endpoints measured in such tests include both structural and functional elements. The experimental design is mostly independent of measurement endpoints and mostly dependent on regulatory endpoints. Endpoints of regulatory concern (e.g., biomass change) can effectively be addressed only from an appropriate experimental design. Hybrid test designs which include both multiple replicates and multiple treatment levels offer, in general, the most appropriate design.