The effect of prenatal cocaine use on neonatal behavior was examined, using the Brazelton Neonatal Behavioral Assessment Scale (BNBAS), in a prospective study of women attending a prenatal clinic. Women were interviewed at the end of each trimester. Term infants were assessed with the BNBAS at day 2 (N = 165) (mean = 35.6 h) and at day 3 (N = 108) (mean = 60.1 h). Women averaged 25 years of age with 12 years of education; 48% were African-American and 20% were married. Women who used cocaine were more likely to use tobacco, alcohol, and marijuana. Regression analyses, with the Lester et al. (22) clusters as outcomes, were used to control for covariates of cocaine use such as other substance use and sociodemographic characteristics. On day 2, first, second, and third trimester cocaine/crack use were significantly related to poorer autonomic stability, second and third trimester use were related to poorer motor maturity and tone, and first and second trimester use were associated with an increased number of abnormal reflexes. These relationships were not present on day 3. These results may be transient expressions of the effect of prenatal cocaine exposure on central nervous system maturation.