The specific goal of this project is to test the differential effectiveness of two treatment programs for pregnant substance-abusing women; residential vs outpatient. Polydrug dependent women will be randomly assigned to one of two treatment modalities, residential or outpatient. The dependent variables to be measured will evaluate pregnancy, neonatal and maternal outcomes; Pattern of drug use in pregnancy fetal growth (birth weight, length and head circumference neonatal neurobehavior complications of labor and delivery maternal post-partum drug use and relapse maternal attitudes and adjustment maternal/infant interaction infant growth and development maternal retention in treatment Analysis of outcome will be based on three major groups of subjects: Group I -Women who are polydrug dependent and receive residential treatment services Group II - Women who are polydrug dependent and receive treatment services in an outpatient program. Group III - A comparison group of drug free women who receive prenatal care and deliver through the services of the same center. This group will provide baseline measures. Data analysis through univariate and multivariate analyses will compare pregnancy outcome, the developmental course of the neonate and the infant and the long term course of the mothers as affected by the type of maternal treatment services received.