The Japanese medaka, Oryzias latipes, embryo-larval assay (ELA) was used to evaluate the toxicity of an industrial effluent prior to, during, and after completion of a toxicity reduction plan (July 1985 to May 1987). The ELA parameters evaluated were embryo lethality, developmental abnormalities (teratogenic and grossly visible lesions), developmental stage delays, survival post-hatching, and histologic evaluation of embryos and fry-exposed fish. Fourteen 24-h composite wastewater samples were evaluated for their toxicity. The effluents were tested as dilutions of the effluent or as a dilution of an XAD-2 resin concentrate, with both giving similar dose response curves. The toxicity reduction plan initiated by the industry involved isolation of toxic waste streams, increased removal of heavy metals, upgrading of the treatment facility, and addition of increasing concentrations of powdered activated carbon (PAC) at 50, 120, and 350 mg/L. The effluent prior to the initiation of the toxicity reduction plan was embryo-toxic, teratogenic, and resulted in histopathologic lesions. There was a shift in the effluent toxicity from being acutely toxic (1985), to resulting in sublethal lesions (1986) and, finally, only delays in embryonic stage development (1986/1987). There was a similar decrease in the occurrence and severity of histological lesions in fish allowed to grow for 30 days. There was a steady decrease in acute toxicity and sublethal effects as the amount of PAC that was added to the system. Concomitantly with the decreased toxicity in the ELA system, there was a decrease in chemical concentrations and Ames bacterial mutagenicity levels. Because there were several steps taken by the industry to reduce its effluents toxicity, the decrease in toxicity cannot be attributed to any single modification. The results from these studies demonstrated the usefulness of the ELA assay in monitoring highly toxic effluents, as well as effluents discharged from facilities that are undergoing toxicity reduction plans.