Two parallel case-control studies were conducted in the Marseilles metropolitan area of France from 1979 to 1985 on cancers and adenomatous polyps of the colorectum. All cases of cancer (n = 389) and polyps (n = 252) were incident and histologically confirmed. Controls (n = 641) matched for sex and age were selected among patients undergoing functional reeducation for injury or trauma in one of five hospital centers. Intake of alcoholic beverages was investigated by questions on daily or weekly intake of wine, beer, apéritifs, and distillates during different life periods. Questions on alcoholic beverages were integrated in a detailed diet history interview questionnaire. The risk of rectal cancer was elevated in male beer drinkers (relative risk = 1.73, 95% confidence interval 1.01-2.95) and in men and women combined (relative risk = 1.71), while beer consumption was not associated with colon cancer. Total ethanol intake and consumption of wine and distillates were not associated with the risk of cancer of the colon or rectum, nor with risk of polyps. Adjustment in the statistical analysis for energy intake and for other dietary variables (fiber from fruits and fiber from vegetables), which were found to be risk factors in the study, did not substantially change the results found for beer and other alcoholic beverages. Etiologic hypotheses related to nitrosamine content of beer are discussed in the light of growing epidemiologic evidence that beer consumption specifically increases the risk of rectal cancer.