Young women with breast cancer have been reported to have an increased risk of dying from their disease if they have given birth in <2 years before diagnosis. The prognostic factors associated with the tumors of these women have not been thoroughly studied. We examined the tumors of the women who had a recent birth and compared the tumor characteristics with those of women who were nulliparous or had given birth > or =5 years before diagnosis. A follow-up study was conducted of 1174 women <45 years old whose invasive ductal breast cancer was diagnosed from January 1983 to December 1992 in three counties of western Washington. These women had participated previously in a population-based, case-control study. Mean follow-up time was 105.4 months. Histological slides were collected for 79.1% of the tumors and reviewed by the study pathologist. Using immunoperoxidase assays, tumor tissue was tested for prognostic markers for 70.4% of the tumors from the women. Cox proportional hazards models were used to estimate the relative risk of dying from breast cancer associated with reproductive events. Logistic regression was used to obtain estimates of the association between various reproductive factors and tumor characteristics. At the end of follow-up, 48.2% of the women (n = 83) whose last birth occurred in < 2 years of diagnosis had died, compared with 23.3% of nulliparous women (n = 189) and 24.4% of the women (n = 661) whose last birth was > or =5 years before diagnosis. The tumors of the women with a recent birth (<2 years before diagnosis) were more likely to be progesterone receptor negative, odds ratio (OR) = 2.2, 95% confidence interval (CI) = 1.2-3.9, to be p53 positive, OR = 2.6, 95% CI = 1.5-4.7, to be of high histological grade, OR = 5.9, 95% CI = 1.7-20.1, to have high mitotic count, OR = 2.2, 95% CI = 1.4-4.4, to be node positive, OR = 2.1, 95% CI = 1.3-3.5, to have a high S phase fraction, OR = 2.3, 95% CI = 1.1-4.8, and to have a high American Joint Committee on Cancer stage (III+), OR = 2.8, 95% CI 1.3-5.8, compared with the tumors of nulliparous women. After adjusting for tumor characteristics and treatment, the risk of mortality associated with a birth in < 2 years of diagnosis of breast cancer remained an independent predictor of mortality, hazard radio (HR) = 2.7, 95% CI = 1.6-4.3. Our study provides evidence that reproductive factors influence the biological behavior of breast cancer in young women and prognosis. Clinicians need to be aware that women who have delivered a child in < 2 years before diagnosis are at increased risk of having tumors with especially adverse prognostic profiles and have a poorer survival rate than women who are nulliparous or whose last birth was some years in the past.