The effects of reproductive factors on the etiology of colon cancer in Asian populations remain unexplored. So we examined 38,420 Japanese women aged 40-79 years who responded to a questionnaire on reproductive and other lifestyle factors from 1988 to 1990 in the Japan Collaborative Cohort Study for Evaluation of Cancer Risk. During an average 7.6 years of follow-up, we documented 207 incident colon cancers. Multivariate analysis indicated that colon cancer risk was likely to be lower among pa-rous women than among nulliparous. Women who had two abortions or more had a 72% higher risk of developing colon cancer [relative risk (RR) 1.72; 95% confidence interval (CI) 1.16-2.55; trend P<0.01] compared with women who never had an abortion. The RR of colon cancer among postmenopausal women significantly decreased with increasing age at menarche (trend P=0.01). No apparent association between colon cancer and gravida, age at first birth, age at menopause, or duration of menstruation was seen. These prospective data support the hypothesis that female reproductive events modify colon cancer risk, and suggest that reproductive factors, particularly age at menarche and having an abortion, may be of importance in the etiology of colon cancer among Japanese women.