Data from four generic microcosm experiments examining the effects of toxic chemicals were used to examine statistical and nonstatistical considerations in quantifying pollutant-induced changes in aquatic microcosms. The minimum limits of detectable change within the experimental design were assessed in terms of mean ecosystem dissolved oxygen production and consumption, and algal community primary productivity rate. Consistent reproducibility in hazard rankings based on the area of “relative impact” was found for response variables in microcosms treated with four toxic chemicals in repeat experiments. Based on changes in microcosm dissolved oxygen production, the response pattern for atrazine treatment was found to be similar among repeat experiments. Primary productivity response data for the four experiments could not be fitted to a single linear or curvilinear regression model, and, therefore, the similarity of repeat test results could not be analyzed statistically.