The association between intake of methylaxnthines and risk of breast cancer was examined in this population-based case-control study conducted in Adelaide, South Australia. The study included 451 cases, and one control was matched to each case for age (to the nearest year). Overall, and within the post-menopausal stratum, there was relatively little variation in the risk of breast cancer in association with total caffeine and total methylxanthine intake. In pre-menopausal women, risk was increased at the higher levels of intake, but the increase in risk was not dose-dependent. There was no evidence for an effect of methylxanthines at high levels of fat intake. Prior epidemiological and experimental studies have not provided strong support for a positive association between intake of methylxanthines and risk of breast cancer. However, the possibility that they are linked through an association between methylxanthines and those histological subcategories of benign breast disease which are strongly associated with risk of breast cancer is discussed.