We completed a phase I trial of indole-3-carbinol (I3C) in 17 women (1 postmenopausal and 16 premenopausal) from a high-risk breast cancer cohort. After a 4-week placebo run-in period, subjects ingested 400 mg I3C daily for 4 weeks followed by a 4-week period of 800 mg I3C daily. These chronic doses were tolerated well by all subjects. Hormonal variables were measured near the end of the placebo and dosing periods, including determination of the urinary 2-hydroxyestrone/16alpha-hydroxyestrone ratio. Measurements were made during the follicular phase for premenopausal women. Serum estradiol, progesterone, luteinizing hormone, follicle-stimulating hormone, and sex hormone binding globulin showed no significant changes in response to I3C. Caffeine was used to probe for cytochrome P450 1A2 (CYP1A2), N-acetyltransferase-2 (NAT-2), and xanthine oxidase. Comparing the results from the placebo and the 800 mg daily dose period, CYP1A2 was elevated by I3C in 94% of the subjects, with a mean increase of 4.1-fold. In subjects with high NAT-2 activities, these were decreased to 11% by I3C administration but not altered if NAT-2 activity was initially low. Xanthine oxidase was not affected. Lymphocyte glutathione S-transferase activity was increased by 69% in response to I3C. The apparent induction of CYP1A2 was mirrored by a 66% increase in the urinary 2-hydroxyestrone/16alpha-hydroxyestrone ratio in response to I3C. The maximal increase was observed with the 400 mg daily dose of I3C, with no further increase found at 800 mg daily. If the ratio of hydroxylated estrone metabolites is a biomarker for chemoprevention, as suggested, then 400 mg I3C daily will elicit a maximal protective effect.