Maternal and neonatal growth, behavior, and physiologic organization were evaluated in 104 mother-infant pairs with positive results of urine toxicology screens. ANOVA comparison of cocaine, methamphetamine, and cocaine plus methamphetamine groups revealed no significant differences in perinatal variables. The Finnegan withdrawal scoring scheme demonstrated that all three groups of infants had altered neonatal behavioral patterns, characterized by abnormal sleep patterns, poor feeding, tremors, and hypertonia. Infants exposed to cocaine or methamphetamine or both were combined and compared with both narcotic-exposed and drug-free mother-infant pairs matched for known maternal risk factors. All drug-exposed groups had significantly higher rates of prematurity and intrauterine growth retardation and smaller head circumferences than did the drug-free comparison group. A significantly higher rate of placental hemorrhage occurred in the cocaine plus methamphetamine group. Stepwise multiple regression analysis assessed the independent contribution of maternal factors; cocaine or methamphetamine was adversely, negatively associated with gestational age, birth weight, length, and occipitofrontal circumference. The increased rate of prematurity, intrauterine growth retardation, and perinatal complications associated with perinatal exposure to cocaine or methamphetamine was greater than that predicted by coexisting risk factors and was consistent with the pharmacologic properties of these drugs.