Dietary pattern analysis is a powerful technique to study the relationships between diet and cancer, accounting for the specificities of overall dietary intake in each setting. The objective of this study was to quantify the association between dietary patterns and colon and rectum cancers. We evaluated 151 rectum and 102 colon cancer cases selected among surgical patients at the Portuguese Oncology Institute, and 879 community controls. Dietary intake was assessed using a validated food frequency questionnaire. Dietary patterns were defined with principal components and cluster analyses. Age, sex, education, total energy intake, and physical activity-adjusted odds ratios (OR) and 95% confidence intervals (95% CIs) were computed. Three dietary patterns were identified: (i) PI: 'healthy,' high consumption of dairy products, wholegrain cereals, vegetables/salads, legumes, fruits and vegetable soup, and low intake of wine; (ii) PII: 'low consumption of milk and foods containing dietary fiber,' low intake of dairy products, vegetables/salads, legumes, and fruits; and (iii) PIII: 'western,' high consumption of red/processed meat, refined cereals, sugar and sweets, potatoes, alcoholic beverages, and low intake of wholegrain cereals and vegetable soup. Compared with PI, the risk of colon cancer was higher among subjects with PII (OR = 2.07; 95% CI: 1.04-4.14) and PIII (OR = 2.35; 95% CI: 1.19-4.64). The OR estimates for rectum cancer were 3.12 (95% CI: 1.74-5.61) and 1.41 (95% CI: 0.75-2.63), respectively. Our results confirm the higher risk of colorectal cancer among subjects with 'western' diets and 'low consumption of milk and foods containing foods containing dietary fiber'. At a local level, these results support public health messages based on the accumulated evidence on the relationship between individual food items/groups and colorectal cancer.