The insecticide endosulfan was evaluated for chronic toxicity and carcinogenicity in long-term feeding studies in both Sprague-Dawley rats and NMRI mice. Dietary concentrations of the test substance were administered to rats at 0, 3, 7.5, 15 and 75 ppm and to mice at 0, 2, 6 and 18 ppm for 24 months each. In the rat study, only the treatment with the highest dose caused a significant reduction of body weight gains of the males and females at 75 ppm. The increased incidence of enlarged kidneys seen at autopsy at 75 ppm in females, and the slightly increased incidence of progressive glomerulonephrosis and slightly increased incidence of renal aneurysms seen histopathologically in the 75 ppm males, make the kidney the target organ in rats. On the basis of these findings a dietary concentration of 15 ppm is considered to be the no-observed-effect level (NOEL) in rats, equivalent to a daily test substance intake of 0.6 mg/kg body weight in males and 0.7 mg/kg body weight in females. In the mouse study, the treatment with 18 ppm caused a significant increase of mortality in the females and a slight (in the first third of the study, significant) reduction of body weight gain in males. Since there were no other substance-related findings, the dietary concentration of 6 ppm is considered to be the NOEL in mice, equivalent to a daily test substance intake of 0.84 mg/kg body weight in males and 0.97 mg/kg body weight in females. An evaluation of all relevant tumour data gained in both studies revealed no differences between control and treated groups. It was concluded, therefore, that endosulfan has no carcinogenic potential.