Breast cancers demonstrate a gradient of responsiveness to endocrine therapy according to hormone receptor status, with tumors positive for both estrogen and progesterone receptors responding most favorably. The authors hypothesized that reproductive risk factors, which are probably mediated by endogenous hormones, would also differ according to receptor status, and that tumors positive for both receptors would exhibit the highest risk. Information on risk factors was obtained from 1,154 breast cancer cases and 21,714 cancer-free controls at the Aichi Cancer Center Hospital, Nagoya, Japan between 1988 and 1992. Receptor status was known for 40% of cases. For a given risk factor, odds ratios with respect to the common control group for breast cancers of differing receptor status were compared using multiple polytomous logistic regression. Risk factors did not differ significantly according to estrogen receptor status. However, age at diagnosis/interview, occupation, age at menarche, menstrual regularity at ages 20-29 years, and cigarette smoking differed significantly or borderline significantly in effect according to progesterone receptor status. Stratification of cases according to joint estrogen and progesterone receptor status indicated that estrogen receptor status did not modify this difference in effect. This study did not support the hypothesis of a gradient of risk for reproductive factors according to hormone receptor status. The authors recommend that the finding that some risk factors differ according to progesterone receptor status should be pursued in further studies.