The relation between coffee, decaffeinated coffee, tea and oral/pharyngeal, and esophageal cancer risk is inadequately quantified. Data were derived from hospital-based case-control studies conducted in Italy and Switzerland. The study on oral/pharyngeal cancer included 749 cases and 1772 controls, and that of esophageal cancer 395 cases and 1066 controls. Multivariate odds ratios (OR) and 95% confidence intervals (CI) were computed. The OR for >3 cups/day of coffee compared with </=1 were 0.6 (95% CI 0.5-0.9) for oral/pharyngeal, and 0.6 (95% CI 0.4-0.9) for esophageal cancer, consistent across strata of age, sex, education and alcohol. The inverse trends in risk were significant. No association emerged with decaffeinated coffee (OR 1.1 for oral/pharyngeal and 0.6 for esophageal cancer) or tea (OR 0.9 for both cancers), consumed in low amounts by these populations. Coffee may decrease the risk of oral/pharyngeal and esophageal cancer.