The present studies determined the influence of dietary supplements of garlic powder (0, 1, 2 or 4%) on 7,12-dimethylbenz[a]anthracene (DMBA)-induced mammary tumors and on the in vivo occurrence of mammary DMBA-DNA adducts in rats. Diets were offered 2 weeks before and 2 weeks following DMBA treatment (25 mg/kg body wt). An additional group was fed the 2% garlic powder diet throughout the 20 week study. Although food intake and weight gain were not influenced, dietary garlic powder supplementation did significantly delay the onset of first tumors (P < 0.01) and did reduce the final mammary tumor incidence (P < 0.01). Consumption of garlic powder also significantly depressed the in vivo binding of DMBA to mammary cell DNA. Binding of both anti- and syn-dihydrodiol epoxides to DNA was depressed in rats fed supplemental garlic powder. The activity of glutathione S-transferase (GST) in mammary and liver tissue from rats fed 2% dietary garlic powder was higher than observed in tissues from rats fed the basal diet. No further increase in GST activity occurred when the dietary garlic content was increased from 2 to 4%. Final mammary tumor incidence was found to correlate positively with total DMBA-DNA binding and the quantities of individual DMBA-DNA adducts. The present studies demonstrate that garlic powder is effective in inhibiting DMBA-induced mammary tumors, possibly by reducing DMBA-DNA binding.