BACKGROUND: Nurses and other hospital workers are exposed to antineoplastic drugs during daily activities. Previous studies suggest that antineoplastic drugs at occupational exposure levels may be toxic to reproduction, but these studies are not consistent or conclusive. METHODS: Self-administered questionnaires were completed by 4393 exposed and nonexposed nurses employed between 1990 and 1997 (79% response). Questions were asked about pregnancy outcome, work-related exposures, and lifestyle. Exposure to antineoplastic drugs was estimated using task-based dermal exposure measurements and self-reported task frequencies. Time to pregnancy was modeled using survival analysis, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for other reproductive outcomes using multiple logistic regression analysis. Associations were further explored by nonparametric regression modeling. RESULTS: Nurses highly exposed to antineoplastic drugs took longer to conceive than referent nurses (adjusted hazard ratio = 0.8; CI = 0.6-0.9). Exposure to antineoplastic drugs was associated with premature delivery (OR per unit increase in ln[exposure] = 1.08; CI = 1.00-1.17) and low birth weight (OR per unit increase in ln[exposure] = 1.11; 1.01-1.21). Penalized smoothed spline plots corroborated these log-linear relations. Spontaneous abortion, stillbirth, congenital anomalies, and sex of offspring appeared not to be related to exposure to antineoplastic drugs. CONCLUSION: Antineoplastic drugs may reduce fertility and increase poor neonatal outcomes among occupationally exposed oncology nurses. © 2007 Lippincott Williams & Wilkins, Inc.