We examined the association between coffee consumption and oral cancer in a hospital-based case-control study comprising 143 patients with oral and oropharyngeal squamous cell carcinoma attended at 3 major hospitals in Sao Paulo, Brazil, and 240 controls without cancer, recruited from outpatient units of the same hospitals and matched with cases by sex and age. Associations were assessed by multivariate logistic regression conditioned on sociodemographic and behavioral characteristics. Tobacco smoking, alcohol drinking, and higher intake of bacon and deep-fried foods were directly related to disease; the inverse was observed to family income and salad intake. Coffee consumption and tobacco smoking were partially correlated (Spearman correlation coefficient 0.14 among cases, 0.31 among controls). When adjusted for all covariates, a cumulative coffee consumption higher than 18.0 daily liters × year during lifetime was indicated to be protective against disease (adjusted odds ratio 0.39, 95% confidence interval 0.16-0.94, P = 0.037). This observation may have pharmacological implications for clinical medication of these cancers and is relevant to programs aimed at reducing the burden of disease.