Mortality rates among 838 white female breast cancer patients were examined for relationships to personal characteristics assessed at time of diagnosis. These included weight and body mass index, oral contraceptive use, and prior menstrual and reproductive events. Multivariate analyses were used to determine the importance of these characteristics in predicting death rates specific for age and extent of disease at time of diagnosis. Weight was associated with poor prognosis among premenopausal patients, but not among postmenopausal patients. Premenopausal women weighing more than 140 lbs at diagnosis had death rates 1.7 times those of lighter women (P = 0.04). This effect was not explained by differences in the distribution of disease stage between the two groups. Body mass index was positively associated with mortality in both premenopausal and postmenopausal patients. In the entire group, the death rate ratio was 1.4 (P = 0.02) for obese (body mass index greater than 30.4 X 10(-3) lbs/in.2) vs lean women. Among premenopausal patients, parous women had higher mortality rates than nulliparous women (death rate ratio = 2.0, P = 0.06). Although the data were sparse, death rates were higher for women having a full-term pregnancy within 2 years of diagnosis than for premenopausal women with earlier pregnancies. Oral contraceptive use, age at menarche, age at first full term birth, and age at menopause were unrelated to mortality.