Background. Insulin-like growth factor (IGF)-I, a mitogenic and antiapoptotic peptide, can affect the proliferation of breast epithelial cells, and is thought to have a role in breast cancer. We hypothesised that high circulating IGF-I concentrations would be associated with an increased risk of breast cancer. Methods. We carried out a nested case-control study within the prospective Nurses' Health Study cohort. Plasma concentrations of IGF-I and IGF binding protein 3 (IGFBP-3) were measured in blood samples collected in 1989-90. We identified 397 women who had a diagnosis of breast cancer after this date and 620 age-matched controls. IGF-I concentrations were compared by logistic regression with adjustment for other breast-cancer risk factors. Findings. There was no association between IGF-I concentrations and breast-cancer risk among the whole study group. In postmenopausal women there was no association between IGF-I concentrations and breast-cancer risk (top vs bottom quintile of IGF-I, relative risk 0.85 [95% CI 0.53-1.39]). The relative risk of breast cancer among premenopausal women by IGF-I concentration (top vs bottom tertile) was 2.33 (1.06-5.16; p for trend 0.08). Among premenopausal women less than 50 years old at the time of blood collection, the relative risk was 4.58 (1.75-12.0; p for trend 0.02). After further adjustment for plasma IGFBP-3 concentrations these relative risks were 2.88 and 7.28, respectively. Interpretation. A positive relation between circulating IGF-I concentration and risk of breast cancer was found among premenopausal but not postmenopausal women. Plasma IGF-I concentrations may be useful in the identification of women at high risk of breast cancer and in the development of risk reduction strategies. Additional larger studies of this association among premenopausal women are needed to provide more precise estimates of effect.