The relationship between smoking habit and risk of colon and rectal cancers was considered in a case–control study conducted between 1991 and 1996 in six Italian centers. Cases were 1225 patients below age 75 with histologically confirmed cancer of the colon and 728 with cancer of the rectum.Controls were 4154 patients admitted to hospital for a wide spectrum of acute, non-neoplastic diseases. Compared to neversmokers, the odds ratios (OR) for current smokers of 25 or more cigarettes/day was 0.90 for patients with colon and 0.86 for those with rectal cancer and those for ex-smokers were 1.02 and 1.09 for colon and rectal cancer, respectively. No increase in risk was found with duration of the habit, the OR for 40 or more years being 0.79 for colon and 0.87 for rectal cancer. Furthermore, no relationship was apparent with time since starting (the OR for 40 or more years were 0.94 for colon and 1.01 for rectal cancer), or age at starting (the OR for < 18 years were 1.02 for colon and 1.00 for rectal cancer), or for pack-years smoked (the OR for 40 or more pack-years were 0.93 for colon and 0.91 for rectal cancer) or time since stopping among ex-smokers. No increase in risk was found in smokers of ≥ 15 cigarettes/day for 40 years or longer (OR: 0.93). No significant heterogeneity was found across strata of age at diagnosis, sex, education, physical activity at work, intake of alcohol, coffee, vegetables, total energy, and number of meals/day. Likewise, no significant association was apparent for various intestinal subsites. Thus, this study did not find cigarette smokers at higher risk for cancer of the bowel even after a long duration and a long period since starting.