Colon cancer is the second most common cancer in both men and women in North America and other developed countries. In a population-based case-control study of colon cancer among French Canadians in greater Montreal, a total of 402 cases and 668 controls were interviewed. The cancer cases were identified through the admission offices of five major Francophone teaching hospitals in Montreal from 1989 to 1993. The controls, matched by age, sex, place of residence, and language, were selected by a modified random digit dialing method. The results show that subjects who had ever been married had a lower risk for colon cancer (odds ratio [OR]: 0.58; 95% confidence interval [95% CI]: 0.48-0.84) than did individuals who had never been married. A significant association (OR: 1.90; p for trend = 0.003) was found between the height of subjects and the risk of colon cancer. The OR for individuals with a family history of colorectal cancer was 2.78 with a p value of 0.01. A direct and significant association (OR: 2.01) was found among constipation, use of laxatives (OR: 1.41), and the risk of colon cancer. Among women, a suggestive inverse association was detected between the number of full-term pregnancies and the risk of colon cancer in female subjects (the OR for five or more pregnancies was 0.58 with a p for trend of 0.08). There was also a suggestive linear trend (increased age-decreased risk) between age at menarche and the risk of colon cancer. No association was apparent between other sociodemographic characteristics and the risk of colon cancer. In conclusion, married individuals had lower risk for colon cancer, perhaps due to food habits or other characteristics of being single. Higher height and weight history 10 years before the diagnosis of cancer are risk factors for breast cancer, while both current weight and body mass index seem to be protective. Positive family history of colon cancer increased the risk of colon cancer significantly.