OBJECTIVE Adverse health effects are associated with intrauterine cocaine exposure. The purpose of this study was to investigate the impact of fetal cocaine and crack exposure on neonatal outcome. METHODS We enrolled 386 mother-infant pairs, including 130 matched control pairs, in the study. The course of pregnancy and delivery was followed and neonatal outcome was assessed by physical and neurologic examination, as well as by the Brazelton Neonatal Behavioral Assessment Scale and the Neonatal Stress Scale. RESULTS The cocaine-exposed neonates had significantly more adverse effects than the matched control infants. Birth weight, length, and head circumference were significantly lower in the cocaine- and crack-exposed infants (p < or = 0.001). There were significantly more premature infants (p < or = 0.007) in this group. They demonstrated significant abnormalities on the neurologic examination (p < or = 0.001), inferior performance on the Brazelton Neonatal Behavioral Assessment Scale (p < or = 0.001), and higher scores on the Neonatal Stress Scale (p < or = 0.001). Predictors of negative neonatal outcome were maternal age (p < or = 0.02), poor paternal relationship with the mother (p < or = 0.002), crack use (p < or = 0.004), cocaine use (p < or = 0.009), and marijuana use (p < or = 0.05). CONCLUSIONS The single most important predictor of neonatal outcome is the frequency, quantity, and type of cocaine used.