A population-based case-control study was conducted between July 1984 and February 1988 in the Spanish island of Majorca; 286 incident colorectal cancer cases, 295 population controls and 203 hospital controls were interviewed using a food frequency questionnaire. In a multivariate analysis, an increased risk of colon cancer was found for high consumption of fresh meats (RR =2.87) while a high consumption of cruciferous vegetables afforded protection (RR =0.48). For rectal cancer an increased risk was associated with dairy products (RR =3.08) while a protection was afforded by consumption of cruciferae (RR =0.50). For colorectal cancer, the cereal food group also showed an increase in risk (RR =1.92). When cases were compared to hospital controls, the effects of cruciferae in colon and rectum and those of dairy products in rectal cancer remained. The magnitude of the RR estimates was decreased for most comparisons, although in general terms the direction of the associations was the same. In addition, univariate analyses of food groups also suggested significant increases in risk of colorectal cancer for increasing consumption of cereals, potatoes, pastry, eggs and number of meals per day. An indication was found of a reduction in risk for consumers of coffee. An analysis based on risk scores was also conducted and a 4-fold increase in the risk of colorectal cancer and a highly significant statistical trend was found for high consumption of fresh meat, dairy products and cereals combined with low consumption of cruciferae.