Maternal substance abuse during pregnancy is a growing problem with major public health and legal concerns. In utero substance exposure may adversely affect neonatal development; pregnancy outcome; and the long-term behavioral, cognitive, and developmental abilities of the child. Also, serious legal implications are associated with substance abuse during pregnancy, including charges of child abuse and neglect that may result in the removal of the neonate from parental care and loss of custodial rights. Timely detection of in utero drug exposure is necessary for early identification and effective management of exposed newborns. Accurate identification of drug-exposed newborns relies on maternal history; clinical presentation of the newborn; and laboratory testing of biological maternal matrices (ie, urine, blood, oral fluid, sweat, hair, and breast milk), neonatal matrices (ie, urine, meconium, hair, and umbilical cord blood and tissue), and/or matrices from both the mother and neonate (ie, placenta and amniotic fluid). Evaluation of biological matrices can account for in utero exposure at various stages of gestation and approximate the period (recent versus chronic use) of substance exposure. Each matrix has its own unique advantages and limitations in terms of ease of collection, the window of gestational exposure represented, and sensitivity for different parent drug analytes and metabolites, which must be carefully considered for accurate interpretation of results. Analytical approaches to sample preparation and analysis vary based on the complexity of these biological matrices. Immunoassays are routinely used for screening, and chromatographic separation coupled to mass spectrometry detection method is commonly used for definitive (confirmatory) testing. Some laboratories use a single technology for all testing. This review provides a discussion on approaches used to detect drug-exposed newborns, biological specimens that have been studied to identify and characterize drug exposures, example analytical methods for meconium and umbilical cord tissue as well as considerations surrounding the interpretation of results. A possible algorithm for testing is also proposed.