The association of beer drinking with cancer of the rectum was investigated in a case-control study of 130 male and 88 female rectal cancer cases and 336 male and 249 female controls. Information was obtained on consumption of beer, wine, and hard liquor throughout adulthood (quantity and duration), as well as on smoking and sociodemographic characteristics. Beer intake was not significantly associated with estimated risk of rectal cancer in females but was in males, with an increasing gradient in the odds ratio (OR) with increasing beer consumption. For drinkers of 32 or more ounces of beer per day, the OR was 3.5 (95% Cl 1.8–7.0). No association was seen with duration of beer drinking. Wine and hard liquor consumption showed no association with the development of rectal cancer. In multiple logistic regression analyses, the relative risk for beer drinking was reduced slightly when potential confounding variables were included in the model (RR adjusted for religion and education: 2.7, 95% Cl 1.3–5.7). The study results are discussed in the light of other epidemiological studies of rectal cancer and beer drinking. We conclude from the aggregate evidence that the association of beer drinking with rectal cancer is probably not causal and that the slightly elevated OR's observed for males in this study are most likely due to incomplete control for confounding variables.