Summary Hormone concentrations in blood and total 12 h urine values were compared between 40 postmenopausal women with breast cancer and 40 control women in a study which carefully controlled for the possible confounding effects of age, weight and pregnancy history by individually matching cases and controls on these factors. Breast cancer cases had received only surgical treatment for their localised disease, which was diagnosed from 1 to 9 years before hormonal evaluation. Cases had 15% higher serum oestradiol levels (P=0.02), 40% more urinary oestradiol (P=0.03) and 44% more urinary oestriol (P=0.04) than control women. Cases also had higher levels of serum and urinary oestrone, but these differences were not statistically significant. The percentages of serum oestradiol not bound to albumin or sex-hormone binding globulin did not differ between cases and controls, nor were there statistically significant differences in the serum levels of prolactin, sex-hormone binding globulin or dehydroepiandrosterone sulphate. These results provide further support for the hypothesis that breast cancer risk is determined in part by post-menopausal serum oestrogen concentration.There is a substantial body of epidemiological, experimental and clinical evidence for a role of oestrogens in the aetiology of female breast cancer (Henderson et al., 1988). In premenopausal women, the major source of endogenous oestrogens is the ovary. Menopause signals a marked decline in the amount of circulating oestrogens and this decline is at least part of the explanation for the decreased risk of breast cancer associated with early menopause (Trichopoulos et al., 1972). Although breast cancer incidence continues to rise after menopause in US and European women, the rate of increase in breast cancer incidence decreases substantially. In post-menopausal women, the major source of oestrogen is the peripheral conversion of androstenedione in fat tissue (Grodin et al., 1973). This offers the most probable explanation for the association of obesity with increased risk of breast cancer in post-menopausal women (Lubin et al., 1985).In women, 40-50% of plasma oestradiol (E2) is bound to sex hormone-binding globulin (SHBG) and all but 2-4% of the remainder is bound to albumin (Anderson, 1974). With the development of relevant laboratory technology, attention has focused on evaluating whether the free, non-protein bound fraction of plasma E2 is higher in women with breast cancer than in other women, as free E2 is considered to be freely diffusible into cells and biologically available to receptors in target tissues (Siiteri, 1980); this may also be true of the non-SHGB bound E2 (Pardridge, 1986). This would further explain the effect of body weight on breast cancer risk in older women as increased weight is not only associated with higher post-menopausal oestrogen levels but also with reduced levels of SHBG and, therefore, with greater 'tissue availability' of oestrogens (Siiteri et al., 1981).A number of studies have compared endogenous oestrogen and ...