If having children is regarded as an exposure in life, its effect on a host could be considered as being due to female sex hormones associated with pregnancy in women and some lifestyle factors associated with large families in both men and women. To explore the roles of having children in the etiology of colon cancer, we examined 36,629 women and 24,877 men aged 40-79 years who completed a questionnaire on the number of children and other lifestyle factors from 1988 to 1990 in the Japan Collaborative Cohort Study for Evaluation of Cancer Risk. During 291,080 and 200,648 person-years of follow-up, we documented 198 female and 202 male incident colon cancers, respectively. After adjusting for some factors known or suspected to modify the risk of colon cancer, compared with the women with no children, the multivariate-adjusted relative risks of colon cancer were 0.74 (95% confidence interval [CI]: 0.30-1.84) for one child, 1.00 (95% CI: 0.46-2.20) for two, 0.70 (95% CI: 0.31-1.55) for three, and 0.59 (95% CI: 0.26-1.33) for four or more. The risk of colon cancer showed a significantly monotonic decrease with increasing number of children (P value for trend = 0.047). There was no association between the number of children and colon cancer risk among men. From these prospective data, having children may reduce risk of colon cancer among women, but not among men, suggesting that modifications of hormone profiles secondary to pregnancies may influence female colon cancer risk.