To examine whether vegetable and fruit intake modify colorectal cancer risk, a case-control study was conducted in Uruguay. Dietary patterns were assessed in detail (for cases before diagnosis or symptoms occurred) by use of a food frequency questionnaire on 61 food items, which allowed the calculation of total energy intake. Nutrient residuals were calculated through regression analysis. After adjustment for potential confounders (which included body mass index, total energy, and total alcohol intake), a reduction in risk for total vegetable intake, total fruit intake, and lettuce, apple, and banana consumption was observed. The strongest protection was observed for banana intake (odds ratio 0.28; 95% confidence level 0.16-0.50) for consumption in the third tertile compared with the first.