Curcumin and quercetin were evaluated in rats for their ability to modulate the carcinogenic activity of azoxymethane (AOM) in the colon and 7,12-dimethylbenz[a]anthracene (DMBA) in the mammary gland. In the AOM-induced colon cancer model, male Fischer 344 rats at 8 weeks of age started to receive either curcumin (8 and 16 g/kg) or quercetin (16.8 and 33.6 g/kg) in the diet and 1 week later, were administered AOM (30 mg/kg body wt.) by subcutaneous injection. The animals continued to receive the two agents in the diet until sacrificed 45 weeks later. Curcumin mediated a dose-dependent inhibition of the incidence and multiplicity of adenomas from 47% and 0.58 +/- 0.12 adenomas/rat in the AOM-treated control group to 19% and 0.22 +/- 0.08 and 0.06% and 0.08 +/- 0.06 adenomas/rat for the low and high dose groups, respectively. A low yield of adenocarcinomas (0.06 +/- 0.04 adenocarcinomas/rat) was induced by AOM which was not significantly altered by curcumin. Treatment with quercetin caused a dose-dependent increase in the yield of AOM-induced tumors in the colon from 0.06 +/- 0.04 adenocarcinoma/rat to 0.64 +/- 0.12 and 1.14 +/- 0.17 for the low and high dose groups, respectively. In the DMBA-induced mammary cancer model, curcumin or quercetin was administered at either 10 or 20 g/kg diet, beginning 7 days prior to DMBA and continually throughout the remainder of the experiment. Neither curcumin nor quercetin significantly altered the incidence of animals with tumors or the tumor multiplicity, while the high concentration of both agents significantly increased tumor latency. These results demonstrate different responses to these agents in the two models. While curcumin was highly effective as a chemopreventive agent in the colon model, it was only weakly effective in the mammary model. In contrast, quercetin which was also only weakly effective in the mammary model, caused a dose-dependent enhancement of tumors induced by AOM in the colon model.