BACKGROUND: Data of epidemiological studies on the relation between coffee drinking and upper aerodigestive tract cancer risk are scattered and inconclusive. We therefore conducted systematic meta-analyses of observational studies published before October 2009. MATERIALS AND METHODS: We combined relative risks (RR) with 95% confidence intervals (CI) for cancers of the oral cavity/pharynx (OP) and larynx, esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC), comparing the highest versus the lowest categories of coffee consumption, using random-effects models. RESULTS: For OP cancer, the pooled RR was 0.64 (95% CI 0.51-0.80) for highest versus lowest coffee drinking, based on a total of 2633 cases from one cohort and eight case-control studies, with no significant heterogeneity across studies. The RRs were 0.61 (95% CI 0.41-0.89) for European, 0.58 (95% CI 0.36-0.94) for American and 0.74 (95% CI 0.48-1.15) for Asian studies, where coffee consumption is lower. The corresponding RRs were 1.56 (95% CI 0.60-4.02) for laryngeal cancer (732 cases from three case-control studies), 0.87 (95% CI 0.65-1.17) for ESCC (2115 cases from one cohort and six case-control studies) and 1.18 (95% CI 0.81-1.71) for EAC (415 cases from three case-control studies). CONCLUSION: Coffee drinking is inversely related to OP cancer risk, while there is no relation with laryngeal cancer, ESCC and EAC.