BACKGROUND AND OBJECTIVES: Many neonatal intensive care units (NICUs) do not give rotavirus vaccines to inpatients due to a theoretical risk of horizontal transmission of vaccine strains. We aimed to determine incidence and clinical significance of vaccine-strain transmission to unvaccinated infants in a NICU that routinely administers pentavalent rotavirus vaccine (RV5). METHODS: This prospective cohort study included all patients admitted to a 100-bed NICU for 1 year. Stool specimens were collected weekly; real-time quantitative reverse-transcription polymerase chain reaction was used to detect any RV5 strain. Incidence of transmission to unvaccinated infants was calculated assuming each unvaccinated patient's stool contributed 1 patient-day at risk for transmission. Investigations and geospatial analyses were conducted for suspected transmission events. RESULTS: Of 1238 infants admitted, 560 (45%) were premature and 322 (26%) had gastrointestinal pathology. During observation, 226 RV5 doses were administered. Overall, 3448 stool samples were tested, including 2252 from 686 unvaccinated patients. Most (681, 99.3%) unvaccinated patients never tested positive for RV5 strain. Five (<1%) tested RV5 strain positive. The estimated rate of transmission to unvaccinated infants was 5/2252 stools or 2.2/1000 patient-days at risk (95% CI: 0.7-5.2). No gastroenteritis symptoms were identified in transmission cases within 7 days of collection of RV5-positive stool. Of 126 patients for whom the RV5 series was initiated before the discharge date, 55% would have become age-ineligible to start the series if vaccination was allowed only at discharge. CONCLUSIONS: Transmission of RV5 strain was infrequent and without clinical consequences. Benefits of allowing vaccine-induced protection against rotavirus disease in infants through in-NICU RV5 vaccination appear to have outweighed risks from vaccine-strain transmission.