In a case-control study nested within the Greek EPIC cohort, serum levels of estrone, estradiol, androstenedione, dehydroepiandosterone sulfate, testosterone and IGF-1 were measured for 29 breast cancer patients and 58 control women, matched for age and menopausal status. There was little difference in breast cancer risk when values of 4-6-as compared to values of 1-3-hormones were elevated, a finding arguing against a positive interaction among these hormones. Breast cancer risk, however, was significantly and substantially lower among women with levels of all hormones below the corresponding age- and menopausal-status-predicted means, compared to women with levels of at least 1 hormone above the predicted mean (odds ratio = 0.11 with 95% confidence interval 0.01-0.90; p = 0.04). Our results suggest that the studied mammotropic hormones may act as permissive factors for breast cancer occurrence, and that the levels of some of them above the mean suffice for sustaining growth of a developing tumor. A corollary is that studies of mammotropic hormones in relation to breast cancer risk may also need to focus on the lower end of the distributions of these growth-enhancing hormones.