A population-based case-control study was conducted in King County, WA, to investigate whether risk factors for estrogen receptor (ER)-rich and ER-poor breast cancers differ. Responses to interviews with 329 women with breast cancer who were between 25 and 54 years of age at the time of diagnosis were compared to responses of 332 women of similar age who were selected from female residents of King County by random digit dialing. Of the 329 interviewed cases, 143 had ER-rich tumors, 97 had ER-poor tumors, and 89 had tumors that were not assayed for receptors. The relative risks of ER-rich and ER-poor breast cancers were similar with respect to late menarche, single marital status, history of extended lactation, menopause before age 40, history of benign breast disease, positive family history of breast cancer, obesity, and history of oral contraceptive and noncontraceptive estrogen use. Late age at first full-term pregnancy was a risk factor for ER-rich breast cancer but not for ER-poor breast cancer. This finding suggests that different causal mechanisms operate for these two types of breast cancer and supports the hypothesis that an early first birth protects against breast cancer by reducing the level of ERs in the mammary epithelial cells from which carcinomas develop.