Prescribing patterns and appropriateness of morphine use in a neonatal intensive care unit (NICU) were evaluated in a concurrent drug-use evaluation (DUE). Data were collected for 99 infants who received morphine over a six-month period. Patient charts were reviewed to collect the following data: patient's age, weight, dosage schedule, concurrent sedatives, ventilatory status, whether adequacy of analgesia was documented, and descriptions of adverse drug reactions (ADRs). The physicians' orders were reviewed to determine whether NICU morphine dosage guidelines were followed and whether the indication for use was noted. Seven ADRs occurred in six of the patients; three of the ADRs occurred after ophthalmic cryosurgery. Indications for use were noted in 79 of 285 physician orders (27.7%). The adequacy of sedation or analgesia was documented on 60 of the 360 patient days (16.7%). The DUE results prompted several changes: physicians were asked to select indications from a list in the computerized order-entry system, an analgesia or sedation assessment scale was added to nursing flow sheets, and endotracheal intubation became a requirement before ophthalmic cryosurgery. A follow-up DUE showed nearly complete compliance with the new guidelines for morphine use and a reduction in the number of adverse reactions to morphine. A DUE prompted policy changes that improved documentation of indications for and efficacy of morphine use and reduced adverse reactions to the drug in an NICU.