Background: The incidence of nosocomial blood stream infections (BSI) among NICU admissions remains high, with significant mortality and morbidity. Due to COVID-19, there are increased infection prevention (IP) measures in NICUs including universal masking for all healthcare workers and families, social distancing, visitation restrictions, and increased attention to hand hygiene. These measures may also affect late-onset infection rates and offer understanding of novel interventions for prevention. Methods: We examined infection rates from three neonatal centers during the 24 months prior to implementation of COVID-19 IP measures (PRE-period) compared to the months after implementation from April 2020 (POST-period). Late-onset infections were defined as culture-confirmed infection of the blood, urine, and other sterile fluids or identification of respiratory viral pathogens. An interrupted time series analysis of infection per 1000 patient days was performed based on a change-point Poisson regression with a lagged dependent variable and the number of patient days used as offsets. Each month was treated as independent with additional analysis using an observation-driven model to account for serial dependence. Results: Multicenter analysis to date included all infants cared for at three centers (Level 3 and 4) from 2018-2020. Monthly BSI rates decreased in the POST-period at the three centers (Figure 1). At all centers actual BSI rate was lower than the expected rate in the POST-period (Figure 2). The combined BSI rate per 1000 patient days was 41% lower compared to the rate prior to implementation (95% CI, 0.42 to 0.84, P = 0.004). In subgroup analysis of BSI by birthweight, during the POST-period there was a 39% reduction in infants < 1000g (P = 0.023), a 44% decrease for 1000-1500g patients (P = 0.292) and a 53% decrease in those > 1500g (0.083). Examining single center data from the University of Virginia through March 2021, there was a 36% decrease in all late-onset infections (BSI, UTI, Viral, and peritonitis) (95% CI, 0.46 to 0.90, P=0.011). Conclusion: In this preliminary analysis, we found a reduction of BSI after the implementation of COVID-19 infection prevention measures. Additionally, there were fewer viral infections, though there were a limited number of episodes. Further analyses of multicenter data and a larger number of patients from all 12 centers of our study network will elucidate the significance of these findings and the role some of these IP measures, such as universal masking, may have in infection prevention in the NICU (Supported in part by Grant Funding from the Gerber Foundation).