Background Food components of a diet are highly related, so that building up dietary patterns may help understand the relationship between chronic diseases and diet, and identify high risk groups that need preventive advice. Aim The aim of this study was to determine dietary patterns associated with the colorectal adenoma-carcinoma pathway. Methods We performed a two-step analysis using first principal component analysis to select the most appropriate food groups, then a hierarchical agglomerative clustering method, in order to determine dietary patterns in 1372 subjects included in a case-control study. Patients with hyperplastic polyps (n = 103), adenomas < 10mm, (n = 154) or larger adenomas (n = 208) were then compared with polyp-free controls (n = 426), and colorectal cancer cases (n = 171) compared with population controls (n = 309) using unconditional logistic regression adjusted on age and gender. Results Cluster analysis determined five food patterns. Cluster 1 identified a low-energy diet; cluster 2 a high-starch, highfat, and low-fruit diet; cluster 3 a high-processed meat, -energy, -alcohol, and -starchy foods diet; cluster 4 a high-fish, -cereals, -honey, -olive oil, -fruit and -vegetables diet; and cluster 5 a high-flour, -sugar, -chocolate, -animal fats, and -eggs diet. Logistic regression identified cluster 1 as significantly associated with risk of small adenomas (OR = 1.7; 95% confidence interval 1.0-2.7), large adenomas (OR = 1.9; 1.2-3.0) and cancers (OR = 1.7; 1.1-2.8) compared with cluster 2. Cluster 4 diet was inversely associated with risk of small adenomas (OR = 0.4; 0.2-1.0). There was no relationship between patterns and risk of hyperplastic polyps. Multiple adjustment decreased the strength of the relationships with cluster 1, which remained significantly associated with adenomas, but not cancer. Conclusion A lowenergy diet appeared as protective all along the adenoma-carcinoma sequence, contrary to a high-energy, high-processed meat and -animal fat diet.