Bluegills, channel catfish, white suckers, and rainbow trout were exposed, for several weeks, to various water concentrations of chlorine and chlorine-ammonia mixtures during the summers of 1985 and 1986. The tests were conducted in one-quarter mile long experimental streams which are located along the Mississippi River near Monticello, Minnesota. All streams were continuously supplied with river water. In the 1985 study, streams were treated with chlorine and in the 1986 study, streams were treated with chlorine or a combination of chlorine and ammonia. For both years, chlorine toxicity to the fish was determined by survival and growth effects as related to the water concentration of total residual chlorine (TRC). TRC concentrations in the chlorine treatments were substantially less toxic than in the chlorine-ammonia treatments. Downstream degradation rates of TRC differed between the chlorine and the chlorine-ammonia treatment streams. TRC diel fluctuations at a given sampling site also differed between the two types of treatments. The chlorine streams followed a predictable pattern whereas the fluctuations in the chlorine-ammonia streams appeared to be random.