Epidemiologic evidence indicates that vitamin D is inversely associated with risk of colon or rectal cancer or both. Using data from a case-control study conducted in Italy between 1992 and 1996, we examined the relation between dietary intake of vitamin D and colon and rectal cancer risk. The study population comprised patients with incident colon cancer (n = 1,225) or rectal cancer (n = 728) and 4,154 hospital controls. Odds ratios (OR) and 95% confidence intervals (CI) according to deciles of vitamin D intake were estimated by multiple logistic regression. In addition, we adjusted for intensity of sunlight exposure through stratification by geographic region of residence, and we computed ORs separately by anatomic subsite within the colon. Adjusted ORs for colon cancer were seen to decrease after the 5th decile of vitamin D intake and reached 0.69 (95% CI = 0.50-0.96) for the 9th and 10th deciles, reflecting a statistically significant inverse trend. The inverse association appeared to be somewhat more pronounced for the proximal than the distal colon and was similar among strata of geographic region and calcium intake. Rectal cancer was unrelated to vitamin D intake in this population. In conclusion, we observed an inverse association between dietary vitamin D intake and colon cancer risk among those with the highest intake levels, which was somewhat unexpected given that these levels were still substantially below the levels considered optimal for colon cancer prevention.