In a medical-record-based prospective cohort study of 12,127 pregnancies consecutively delivered during a 52-mo period at a single institution, 204 (1.7%) were complicated by maternal alcohol abuse. A profile of these patients' concomitant risk factors and their pregnancy outcomes indicated that they tended to be older multigravidas who were not currently married. Their obstetric histories were marked by excesses of previous spontaneous abortions, low-birth-weight infants, and fetal anomalies. They were more likely to smoke cigarettes and abuse other drugs. During labor, their risks of infection and of premature placental separation were increased. Evidence of fetal distress during labor and neonatal depression were more common in their infants. Infant birth weights were lower in the alcohol-abusing group by an average of approximately 190 g. This was accounted for by intrauterine growth retardation, which was increased 2.7-fold overall; preterm delivery was not more common. The risk of intrauterine growth retardation was estimated to be increased 2.4-fold in association with alcohol abuse alone, 1.8-fold with smoking alone, and 3.9-fold with these risks together. A significant increase in congenital anomalies (38%) was also identified, though there were no differences between the mortality rates or placental pathology of the alcohol-abusing and comparison groups. These adverse infant outcomes did not appear to be related to demographic factors, the medical care, or nutritional status of the alcohol-abusing patients. Moreover, cigarette smoking did not appear to account for the congenital anomalies observed in the alcohol-abusing group. The findings of this study suggest that alcohol abuse during pregnancy constitutes a significant risk for a range of adverse perinatal outcomes in as many as 50% of the offspring, and hence, should be considered a major public health concern.