The effect of caffeine and/or coffee consumption (via the drinking water) during the initiation phase and promotion phase of 7,12-dimethylbenz(a)anthracene (DMBA)-induced mammary gland tumorigenesis in female Sprague-Dawley rats fed a commercial laboratory animal chow was examined. In the initiation studies, DMBA was administered once at 53-55 days of age; caffeine (100-860 mg/liter of drinking water) and/or coffee (moderate or high dose, sole source of drinking water) treatments were for 32 consecutive days, commencing 29 days prior to DMBA treatment and terminating 3 days after DMBA treatment. In the promotion studies, DMBA was administered once at 54-55 days of age; caffeine and/or coffee treatments were daily from 57-58 days of age to termination of experiments (12-21 weeks after carcinogen treatment). In the initiation studies, either moderate (100-400 mg) or high (860 mg) dose levels of caffeine or moderate to high dose levels of caffeinated coffee significantly (P less than 0.05) reduced mammary carcinoma multiplicity (number of tumors/rat). Consumption of high or moderate dose levels of decaffeinated coffee did not significantly alter mammary carcinoma multiplicity. The addition of caffeine to the moderate dose level of decaffeinated coffee resulted in a significant (P less than 0.05) reduction in mammary carcinoma multiplicity. In the promotion studies, prolonged consumption of moderated dose levels of caffeine or moderate or high dose levels of caffeinated coffee or decaffeinated coffee did not significantly effect mammary carcinoma multiplicity. In the early stages of promotion, however, a significant (p less than 0.05) stimulatory effect of caffeine on mammary carcinoma multiplicity was observed; an effect that was temperate and transitory. In both the initiation and promotion studies caffeine and/or coffee consumption did not significantly affect the incidence of mammary carcinomas (percentage of rats bearing mammary carcinomas) or the mean latency period of mammary tumor appearance. These results provide evidence that caffeine and/or caffeinated coffee consumption can significantly influence mammary carcinoma multiplicity in female rats treated with DMBA, an effect that is dependent upon the dose level, duration, and time-span of caffeine administration.