This study examined the effects of maternal cocaine use on performance on the Neonatal Behavioral Assessment Scale (NBAS). Cocaine-exposed newborns (n = 56) were compared with a non-cocaine-exposed group (n = 30) born to mothers with similar sociodemographic characteristics. Cocaine-exposed newborns showed significant reduction in birth weight but did not experience greater obstetric or postnatal complications. On neurobehavioral assessments using the NBAS, cocaine-exposed newborns showed significantly depressed performance on the habituation cluster but not on other NBAS clusters when differences in birth weight were controlled. In a sample of 30 cocaine-exposed newborns matched on birth weight, gestational age, and race to the 30 non-cocaine-exposed newborns, cocaine-exposed newborns continued to show depressed habituation performance. The significance of a selective effect of cocaine exposure on early habituation performance is discussed in terms of the implications for attentional regulation in the first year of life.