Using data of a case-control study conducted in Italy between 1992 and 1996 on 1225 cases of colon cancer and 4154 controls, we have evaluated the effect of selected established risk factors in subjects with, and in those without, familial propensity (134 cases, 146 controls); there were non-significant associations between education and meal frequency and colon cancer risk. Physical activity, high energy and low vegetable intake were significantly related to colon cancer risk in subjects without familial predisposition, but showed no relationship in those with family history of colorectal cancer: the ORs were 1.1 for the lowest level of physical activity, 0.7 for the highest tertile of energy intake, and 1.0 for the lowest one of vegetable intake. These findings would suggest that genetic predisposition can make the influence of environmental factors for defining the risk of colon cancer weak and/or difficult to estimate.