The relation between dietary factors and the risk of colorectal cancer was investigated in a case-control study conducted in Northern Italy on 339 cases of colon cancer, 236 cases of rectal cancer and 778 controls admitted to hospital for acute, non-neoplastic or digestive disorders. Consistent positive associations were observed with more frequent consumption of starchy foods (pasta or rice) (relative risk, RR = 3.0 for colon and 1.8 for rectum for highest vs. lowest tertile) and beef/veal meats (RR = 2.1 for colon, 2.3 for rectum), whereas reduced relative risks were observed in subjects reporting more frequent green vegetable consumption (RR = 0.5 for highest vs. lowest tertile), a few specific vegetable or fruit items, and coffee (RR = 0.6 for highest vs. lowest tertile). Various fats in seasonings were positively, but inconsistently, related to intestinal cancer risk, whereas no association was evident with measures of whole grain foods or alcohol intake. For both intestinal sites, a 4- to 5-fold difference in risk was evident between the extreme quintiles of a simple score obtained by algebraic sum of the 4 major groups of foods. These findings could not be explained in terms of confounding by socio-economic status or other major potential distorting factors, are in agreement with the results from previous studies of colo-rectal cancer in Southern Europe, and are consistent with various aspects of the descriptive epidemiology of intestinal cancer in Italy.