Manipulations of freshwater sediment were performed to remove indigenous organisms prior to conducting toxicity tests with three species of benthic invertebrates. The effects of these treatments on end points in bioassays were compared within and between two sediments, i.e., a ''clean'' sediment and a ''contaminated'' sediment. In addition, the effects of manipulations on the physicochemical structure of the two sediments and the presence of metals, PAHs, and PCBs in the contaminated sediment were examined. The amphipod Hyalella azteca was most sensitive to the manipulations and had low survival in sediment that was sterilized. Growth (milligrams dry weight per individual) was affected by the presence of contaminants. Survival of Chironomus riparius was not affected by any manipulation but was reduced by contaminants as well as indigenous organisms. Growth of C. riparius was higher in autoclaved sediment but lower in sediment containing endemic tubificid worms. Production of young by Tubifex tubifex increased in sediment that was irradiated, possibly due to increased detrital material. Particle size distribution, metals, nutrients, and PAHs varied little as a function of manipulation; however, sieving of sediment through 250-mu m mesh did reduce percent total organic carbon (TOC), percent loss on ignition (LOI), and concentrations of some PCBs in either clean or contaminated sediment. Manipulation of sediments to remove endemic species should be determined on a case-by-case basis and is specific to the organisms used in toxicity tests.