To evaluate the influence of inhibitors of prostaglandin synthesis on the incidence of DMH-induced colon cancer, 90 male Sprague-Dawley rats were randomly assigned to: indomethacin 20 mg per liter drinking water, meclofenamate 50 mg per liter drinking water, or normal drinking water (control group). Dimethylhydrazine was given by weekly subcutaneous injections (20 mg/kg body weight) during the first 20 weeks. Thirty-two weeks after the start of treatment and carcinogen exposure, the animals were killed and examined for the number, size, location, and spread of intestinal tumors. Colon cancer incidence was significantly lower in animals receiving indomethacin (56 per cent) compared with the control group (88 per cent) and with the meclofenamate group (90 per cent) (P less than 0.005). The corresponding figures for tumors in the small intestine were 31, 46, and 35 per cent, respectively. The tumors in indomethacin-treated animals did not differ in number, size, location, or spread from tumors of the other groups, suggesting that indomethacin might influence the carcinogenic process itself, rather than the natural course of the established disease. We conclude that indomethacin significantly reduces the incidence of large-bowel cancer in this animal model and that this observation may have some potential for future chemopreventive studies in human high-risk groups (e.g. ulcerative colitis, familial polyposis).