To assess whether recent coffee consumption is related to the incidence of large bowel cancer, the authors analyzed data from a multipurpose hospital-based case-control study of several cancers, conducted from 1978 to 1986 in Boston, Massachusetts, New York, New York, Philadelphia, Pennsylvania, and Baltimore, Maryland: 717 cases of colon cancer and 538 cases of rectal cancer were compared with 3,883 controls admitted for nonmalignant and malignant illnesses unrelated to coffee use. Relative risks were estimated by multiple logistic regression with allowance for age, sex, cigarette smoking, alcohol consumption, and several other potential confounding factors. Overall, relative to the consumption of one cup of coffee per day, the estimated relative risk of colon cancer for consumption of up to four cups per day in the period before admission was close to 1.0; for consumption of five or more cups per day, it was 0.6 (95% confidence interval 0.4-0.8). The results were consistent in men and women, and similar results were obtained whether the measure of consumption was use in the year before admission or use three years before admission. For rectal cancer, some relative risk estimates were elevated, particularly in men, but there was no consistent pattern of association, nor was there a trend of increasing risk with increasing consumption. The data suggest that recent coffee consumption is not related to an increased risks of large bowel cancer and that heavy coffee consumption may reduce the risk of colon cancer. A biologic rationale for such a reduction has not been established.