Prenatal cocaine (CC) exposure may result in increased fetal loss, growth retardation, altered neurodevelopment, and sudden infant death syndrome (SIDS). We sought to establish an animal model for prenatal cocaine exposure which (1) would allow us to distinguish the direct effects from the indirect and nutritional effects of the drug, and (2) might be used to address questions of cocaine's toxicity, specifically to the developing respiratory control system. The study design included 38 New Zealand White rabbit does among CC, pair-fed (PF), and free-fed (FF) groups. Miniosmotic pumps were implanted in each doe on day 10 of timed gestation providing continuous subcutaneous administration of either 30 mg/kg/day of cocaine HCl in H2O (CC) or sterile H2O alone (PF and FF). Mean (SEM) plasma cocaine concentration was 1.71 +/- 0.21 mumol/l (519.4 +/- 64.4 ng/ml). Pregnancy outcome compared for incidence of stillbirth, maternal death, spontaneous abortion, and gross malformation among 211 pups was significant only for increased stillbirths among CC pups (18%, p less than 0.04) as compared to PF (6%) and FF pups (7%). External and renal malformation and postnatal weight, crown-rump length, and snout-occiput head circumference for pups aged 4 and 5 days of age did not differ among groups. The direct effects of prenatal cocaine evaluated in our model do not reproduce the altered perinatal outcome observed among humans. However, our results do not determine if physiologic function has been altered. Investigation of the physiologic and pathologic abnormalities that are relevant to this human condition, specifically to the developing respiratory control system, should add clarity to the mechanism of action of cocaine during pregnancy.