Paregoric and phenobarbital, administered randomly in 153 passively addicted neonates, initially appeared to control neonatal abstinence signs equally well. However, seven of the 62 phenobarbital-treated newborns had abstinence-associated seizures within the first month of life, while none of 49 paregoric-treated neonates had seizures. Forty-two neonates initially requiring no specific pharmacotherapy for abstinence signs were born to mothers taking less methadone hydrochloride just before delivery. Five of those 42 neonates, however, had seizures within the first 14 days of life. Seizure occurrence could not be predicted from analysis of early abstinence patterns. We consider paregoric to be the treatment of choice for the neonatal abstinence syndrome. Phenobarbital use should be monitored with serum drug levels and modification of recommended dosage regimens considered.