There is increasing evidence that different types of breast cancers are related to distinct risk factors. We analyzed the risk of breast cancer with respect to circulating insulin-like growth factor (IGF)-I, IGF-binding protein (IGFBP)-3, 17beta-estradiol, estrone, testosterone, androstenedione and sex hormone-binding globulin (SHBG), taking into consideration the characteristics of the tumors. Plasma hormone levels of 102 postmenopausal patients with breast cancer detected by mammography screening, and 102 matched controls were analyzed in relation to the histological type, the status of the estrogen receptor (ER), the progesterone receptor (PR) and the HER2 in the tumors. Significant positive associations were revealed between the IGF-I concentration and the overall risk of breast cancer (OR=3.1, 95% CI: 1.5-6.2), ER+PR+ breast cancer (OR=2.4, 95% CI: 1.1-5.4) and ER+PR- breast cancer (OR=4.3, 95% CI: 1.2-14.3) when the highest and the lowest ranges of IGF-I were compared. Significant associations were also found between the highest and the lowest quartiles of testosterone, resulting in OR=4.1 (95% CI: 1.8-9.4) for the risks of breast cancer and OR=5.8 (96% CI: 2.1-16.2) of ER+PR+ breast cancer. A synergy was seen between IGF-I and testosterone levels. When both plasma IGF-I and testosterone were in the highest quartile ranges, an OR=26.4 (95% CI: 1.6-426.5, p=0.021) was computed for breast cancer overall. No significant synergistic effects could be demonstrated with other parameters. There were significant, 2.5-fold (95% CI: 1.2-5.6), and 16-fold (95% CI: 2.0-133.5) increases in the overall risks of breast cancer and of ER+PR- breast cancer, respectively, when the highest and the lowest quartiles of IGFBP-3 were compared. No associations were found between any of the hormones and the risk of ER-PR- tumors. The increased prevalence of ER+ breast cancers in patients with higher levels of IGF-I, IGFBP-3 or testosterone implicate these hormones in the etiology of hormone-dependent breast cancer. Additional analyses specific for breast cancer subtypes may shed light on the value of hormone determinations for tailored chemoprevention.