A case-control study was conducted in Utah between July 1979 and June 1983 in which 231 cases of colon cancer identified through the Utah Cancer Registry and 391 controls identified through random digit dialing were interviewed. Odds ratios (OR) were calculated comparing the highest exposure categories with the lowest exposure categories. The highest quintile of body mass index (weight (kg)/height (m)2 for males; weight (kg)/height (m)1.5 for females) was associated with increased risk in both males (OR = 2.1) and females (OR = 2.3). In females, total dietary fat (OR = 1.9) and energy intake (OR = 1.5) were associated with an increased colon cancer risk after adjusting for age, body mass index, and crude fiber. Fiber was protective in females (OR = 0.5) after adjusting for age, body mass index, and energy intake, as was beta-carotene (OR = 0.5) after also adjusting for crude fiber. Adjusted risk estimates in males were 2.0 for total dietary fat, 3.8 for polyunsaturated fat, 2.1 for monounsaturated fat, 2.1 for energy intake, 2.5 for protein, 0.3 for fiber, 0.4 for beta-carotene, and 0.3 for cruciferous vegetables. Risk estimates differed by site of cancer within the colon. In males, protein (OR = 3.8) was a risk factor for cancer of the descending colon, while fats (OR = 2.7-8.8) increased the risk of cancer of the ascending colon. The hypotheses that dietary fat increases colon cancer risk while dietary fiber decreases colon cancer risk and that fat and protein may be independently associated with colon cancer risk are supported.