The association of dietary fat with breast cancer in prospective cohort studies has generally been weak and not statistically significant. However, these studies have not considered whether the risk related to fat intake may differ according to estrogen or progesterone receptor status. Dietary habits and other breast cancer risk factors were assessed by mailed questionnaire in January 1986 in 34,388 postmenopausal Iowa women. Through 1991, 724 incident breast cancer cases were ascertained in this cohort using the Iowa cancer registry. Joint estrogen and progesterone receptor status was determined for 479 (66%) breast cancers. For tumors that were positive for both estrogen and progesterone receptors (ER+/PR+) (n = 329), age- and energy-adjusted relative risks for breast cancer adjusted from lowest to highest third of fat intake were 1.0, 1.05, and 1.22 (P trend = 0.14). Corresponding risks for ER+/PR- tumors (n = 75) were 1.0, 0.85, and 1.05 (P trend = 0.86) and for ER-/PR- tumors (n = 61) were 1.0, 1.06, and 0.73 (P trend = 0.68). Only 14 cases were classified as having ER-/PR+ tumors. Adjustment for other breast cancer risk factors did not appreciably alter these findings. There was a suggestion that dietary fat may be associated with ER+/PR+ breast cancers and not other breast cancers. These results are also consistent with an interpretation of no association between dietary fat with breast cancer, regardless of hormone receptor status. It has been suggested that etiological studies of breast cancer should investigate associations according to receptor status. This study provides evidence of a subset of breast cancers that may be related to dietary factors.(ABSTRACT TRUNCATED AT 250 WORDS)