We evaluated whether our previous reports of increased postmenopausal breast cancer risk with higher body mass index (BMI) or of reduced premenopausal and postmenopausal breast cancer risk with higher physical activity levels varied according to the tumor's estrogen receptor (ER) and progesterone receptor (PR) status. Participants enrolled in either of two population-based case-control studies in Los Angeles County, California: one of premenopausal women (ages < or = 40 years), and one of postmenopausal women (ages 55-64 years). Case participants were diagnosed for the first time with in situ or invasive breast cancer from 7/1/83 through 12/31/88 (premenopausal women) or from 3/1/87 through 12/31/89 (postmenopausal women). Joint ER/PR status was collected for 424 premenopausal and 760 postmenopausal case participants. The analysis included 714 premenopausal and 1091 postmenopausal age-matched, race-matched (white or Hispanic), parity-matched (premenopausal women only), and residential neighborhood-matched control participants. Among the postmenopausal women, obesity was associated with an increased odds of ER+/PR+ breast cancer (odds ratio, 2.45 for women in the highest versus the lowest body mass index quartile; 95% confidence interval, 1.73-3.47). Body mass index was associated with neither ER-/PR- tumors among the postmenopausal women nor with any ER/PR subgroup among the premenopausal women. For both premenopausal and postmenopausal women, higher recreational physical activity levels (> or = 17.6 MET-hours/week versus no activity) were associated with a 30-60% reduction in risk of nearly all ER/PR subtypes, although the associations were generally of borderline statistical significance. Examining these potentially modifiable breast cancer risk factors by tumor ER and PR status may provide us with greater insight into breast cancer etiology and the mechanisms underlying the risk factor associations.