Epidemiologic studies have established that increased body weight is associated with a greater incidence of breast cancer in postmenopausal women. The authors hypothesized that abdominal adiposity further increases risk of postmenopausal breast cancer. They therefore investigated the waist-to-hip circumference ratio in relation to breast cancer incidence in a nested case-control study of 41,837 postmenopausal Iowa women aged 55-69 years. Women were recruited through a mail survey in January 1986 and were asked to have someone measure their body circumferences using a paper tape measure and written instructions. Cancer incidence was ascertained using a statewide cancer registry. A total of 229 incident breast cancers occurred among at-risk women during the first 2 years of follow-up (1986-1987). Compared with randomly selected controls (n = 1,839), women with incident breast cancer had a higher age-adjusted mean waist-to-hip ratio (by 0.013 units, p = 0.030), as well as greater mean weight (by 1.7 kg, p = 0.07), body mass index (by 0.6 kg/m2, p = 0.08), and weight gain since age 18 (by 2.7 kg, p less than 0.01). In multiple logistic regression models, age and current body mass index were significant effect modifiers of the association between the waist-to-hip ratio and breast cancer. A two-standard deviation increase in the waist-to-hip ratio (0.168 units) was associated with no increase in the relative risk of breast cancer in younger and lighter postmenopausal women. However, in older, heavier postmenopausal women, the same increase in the waist-to-hip ratio carried greater than a twofold excess relative risk. Adjustment for other breast cancer risk factors did not materially alter this finding. Two plausible explanations for elevated breast cancer incidence in women with abdominal adiposity include 1) increased concentrations of non-protein-bound estrogens due to reduced sex hormone binding globulin, or 2) increased conversion of adrenal androgens to estrone with abdominal adiposity.