Clinical and epidemiological studies on cancer etiology seldom treat coffee drinking as a potential effect modifier. Yet caffeine exerts significant effects upon a large variety of physiologic, cellular and molecular systems. Caffeine, 'the world's most popular drug', is also a fundamental research tool, widely used in clinical studies on drug metabolism, and in experimental studies on cell cycle checkpoints, DNA repair, and apoptosis, among many other. Caffeine can profoundly alter cell cycle checkpoint function and several mechanisms of DNA repair, as well as carcinogen metabolism. The impact of caffeine on cell cycle checkpoint function occurs in spite of it being nonmutagenic in traditional mutagenesis assays. A complex body of biologic evidence suggests that caffeine-containing beverages can both enhance and antagonise potentially carcinogenic exposures. However, most pathways leading to the ultimate effects in human beings remain unknown. It is unclear whether any of the hundreds of compounds contained in coffee and tea exert a direct and significant carcinogenic effect per se in any human tissue at usual conditions of use. Reasons exist to consider that coffee may sometimes be an indirect, positive confounder. The study of interactions between caffeine-containing beverages and environmental agents in well defined groups of healthy and diseased people could yield new insights into checkpoint signal transduction and other mechanisms of carcinogenesis. Information on the use of caffeine-containing beverages should more often be integrated in studies on the role of gene-environment interactions in the pathogenesis of cancer.