These studies were divided into two parts: (1) "time until death" curves in which Paramecium caudatum Ehrenberg was exposed to selected concentrations of the detergent-enzyme Axion, and (2) exposure of fresh-water protozoan communities to selected concentrations of Axion. The former were carried out in test tubes, the latter in plastic troughs with a constantly flowing fluid. Both types of bioassays were carried out with filtered water from Douglas Lake, Michigan, U.S.A. at a temperature of about 25 C. One hundred percent of an exposed population of Paramecium caudatum was killed in 20 to 30 minutes at a concentration of 100 ppm of Axion. No appreciable change in the controls was noted during these tests. The bioassays with fresh-water protozoan communities were carried out in plastic troughs with a constant flow of Douglas Lake water. During the exposure period the lake water flow was stopped and a flow of Axion solution was substituted for approximately three hours. After this exposure period lake water flow were restored. The number of species was determined before and at intervals following exposure. Appropriate controls were maintained. A three hour exposure to 56 ppm (introduced concentration) of Axion caused a 35% reduction in the number of species in less than five hours with recovery to the original number in approximately 145 hours; a three hour exposure to 75 ppm caused a 54% reduction in number of species in less than five hours with recovery to the original number in approximately 240 hours; 100 ppm produced a 55% reduction with recovery in approx. 200 hours; 125 ppm produced a 48% reduction with recovery in approx. 148 hours; 200 ppm caused a 78% reduction with recovery in about 140 hours. Response patterns vary as do those of bioassays with single species. However, it is important to note that each test community had different types of species present, and it is quite likely that some communities had a higher percentage of sensitive species than others. Since this variation also occurs in nature the tests are probably representative of the differences that exist when natural communities are exposed to wastes. Variation in control troughs was less than +- 10% in the number of species.