Metabolism of estradiol occurs via two mutually exclusive hydroxylative pathways, yielding metabolites of divergent biological properties. 2-hydroxyestrone (2OHE1) is anti-estrogenic while 16 alpha-hydroxyestrone (16 alpha OHE1) is a potent estrogen. The ratio of 2OHE1 to 16 alpha OHE1 (2/16 alpha-OHE1 ratio) represents the net in vivo estrogenic activity. In this study, we sought to determine if the urinary 2/16 alpha-OHE1 ratio could be a predictor of breast cancer risk and the factors which influence this ratio. Variables analysed included age at diagnosis, menopausal status, parity, use of oral contraceptives, body mass index, serum levels of insulin-like growth factor-I (IGF-I), IGF binding proteins (BPs) and the presence of breast cancer. Serum and urine were collected from 65 breast cancer patients and 36 controls after an overnight fast. Urinary estrogen metabolites were measured by enzyme immunoassays while serum levels of IGF-I, BP-1 and BP-3 were determined by immunoradiometric assays. 2OHE1 levels and 2/16 alpha-OHE1 ratios were significantly lower (P < 0.05) while 16 alpha OHE1 levels were higher (P < 0.01) in cancer patients. Multiple linear regression analysis showed that levels of urinary metabolites were influenced by parity and breast carcinoma. 2/16 alpha-OHE1 ratio correlated positively with serum BP-3 level (P = 0.03). By multiple logistic regression, 2/16 alpha-OHE1 ratio was the most significant factor predictive of breast cancer. The odds ratio for women with higher 2/16 alpha-OHE1 ratios was 0.10 (0.03-0.38, 95% confidence interval). In conclusion, the profile of urinary estradiol metabolites was distinctly altered in breast cancer patients. In addition, BP-3 may be a potential mechanism by which estradiol metabolites influence breast cancer progression. As 16 alpha OHE1 has been shown to initiate neoplastic transformation of mammary epithelial cells, the 2/16 alpha-OHE1 ratio may serve as a biomarker of increased risk of breast cancer.