OBJECTIVE: The objective of the study was to identify maternal variables predicting length of treatment for neonatal abstinence syndrome (NAS). STUDY DESIGN: This was a retrospective cohort study of infants treated for NAS during 2000-2006 whose mothers were on methadone maintenance at delivery. Mixed-effects linear regression was used to examine the interaction of maternal and neonatal variables with length of treatment. RESULTS: Of 204 neonates born to methadone exposed mothers, the average dose at delivery was 127 mg daily (25-340 mg) with median length of treatment 32 days (1-122 days). Trimester of initial exposure (P = .33), methadone dose at delivery (P = .198), body mass index (P = .31), antidepressant use (P = .40), cigarette use (P = .76), race (P = .78), and maternal age (P = .84) did not predict length of treatment. In the multivariate analysis, gestational age at delivery and benzodiazepine use were significant predictors of length of treatment. CONCLUSION: Later gestational age and concomitant benzodiazepine use were associated with longer treatment.