The effect of diet diversity on colon cancer risk was examined among 205 male and 223 female cases with incident primary histologically confirmed colon cancer and age-, sex-, and neighborhood-matched controls. Diversity was defined as the number of food items on the food frequency interview reported eaten, more than monthly for total foods and for fruits, vegetables, grains, and meats. Adjusted risk of colon cancer associated with total diversity was increased for men in the highest quartile compared with those in the lowest [odds ratio (OR) 1.99, 95% confidence interval (CI) 0.95-4.15] but not for women. There was little association between risk and diversity within specific food groups for either sex, except risk of colon cancer was positively related to meat diversity in men (OR 6.78, 95% CI 2.80-16.45, highest quartile referent to lowest). We found that total diversity was positively related to colon cancer risk independent of other possible confounders. Diversity measures may capture additional diet-related disease risk elements, thus having implications for future recommendations regarding diet and disease.