It has been proposed that breast cancers may differ in their pathogenesis and etiology according to their estrogen receptor (ER) and progesterone receptor (PR) status. This hospital-based case-control study in Japan assessed the relationship between known and suspected breast-cancer risk factors and ER and PR status. Information on risk factors was collected from histologically confirmed breast-cancer cases (n = 519) and from cancer-free controls (n = 9,506). Of 160 cases with known ER status, 58 percent were ER-positive; 38 percent of 157 cases with known PR status were PR-positive. No statistically significant differences were found between ER-positive cf ER-negative cases. However, statistically significant differences between PR-positive cf PR-negative cases were observed for number of full-term pregnancies (P = 0.01), menstrual regularity as a teenager (P = 0.024), and occupation as housewife (P = 0.036). Borderline differences were observed for age at menopause (P = 0.074), and age at menarche (P = 0.083). This study provides some evidence that etiologic distinctions may be greater between PR-positive and PR-negative breast cancers than between ER-positive and ER-negative breast cancers.