We studied the effects of family history on breast cancer risk among 2,908 cases and 3,180 controls, selected from participants in a nationwide screening project. First-degree family history was associated with a twofold risk increase. Second-degree family history effects were minimal, after adjusting for effects of first-degree relatives. Family history effects were not confounded by age at menarche, age at first full-term birth, age at natural menopause, or previous benign breast disease. Risks from mother's and sister's history were independent. The odds ratio (OR) from a maternal history, 1.9 (95% confidence interval [CI]: 1.6-2.3), varied little by the subject's age at diagnosis, menopause status, or disease laterality. Interactions of maternal history effects with multiple breast biopsies and age at menopause were greater than additive, indicating common mechanistic pathways. The OR from a sister's history was 2.3 (95% CI: 1.9-2.8) and was increased among women who were less than 45 (OR = 6.9), had bilateral disease (OR = 4.7), or were premenopausal (OR = 4.4). The effects from a mother's history and a sister's history are modified in different directions by different factors, providing further indication of the separate roles of a mother's and sister's history in breast cancer etiology.