Intrauterine exposure to cocaine results in a variety of perinatal and long-term neurobehavioral sequlae. Maternal history and neonatal urinalysis miss most instances of exposure. After establishing the high sensitivity, specificity, and predictive value of neonatal hair testing for cocaine, the authors embarked on a study to evaluate the clinical use of the test. Between October 1991 and September 1997, they assessed the presence of cocaine in 509 hair samples; 422 were from neonates and were based on clinical suspicion of intrauterine exposure, although history and urine test results were negative. Thirty-two percent were positive, a rate fivefold higher than the incidence found by the authors in population-based studies. Neonates referred by children's aid workers had significantly higher percentages of positivity (60%). Because neonatal hair grows only during the last 4 months of pregnancy, positive neonatal hair test results reflect maternal addiction, and this warrants close follow-up of the baby after release because of the high rate of postnatal risks secondary to suboptimal parental care and because of the proven perinatal risks inflicted by the drug itself.