OBJECTIVE: The aim of this study was to compare outcomes of infants pre and post initiation of a feeding protocol providing an exclusive human milk-based diet (HUM). MATERIALS AND METHODS: In a multicenter retrospective cohort study, infants with a birth weight <1,250 g who received a bovine-based diet (BOV) of mother's own milk fortified with bovine fortifier and/or preterm formula were compared to infants who received a newly introduced HUM feeding protocol. Infants were excluded if they had major congenital anomalies or died in the first 12 hours of life. Data were collected 2-3 years prior to and after introduction of an exclusive HUM diet. Primary outcomes were necrotizing enterocolitis (NEC) and mortality. Secondary outcomes included late-onset sepsis, retinopathy of prematurity (ROP), and bronchopulmonary dysplasia (BPD). RESULTS: A total of 1,587 infants were included from four centers in Texas, Illinois, Florida, and California. There were no differences in baseline demographics or growth of infants. The HUM group had significantly lower incidence of proven NEC (16.7% versus 6.9%, p < 0.00001), mortality (17.2% versus 13.6%, p = 0.04), late-onset sepsis (30.3% versus 19.0%, p < 0.00001), ROP (9% versus 5.2%, p = 0.003), and BPD (56.3% versus 47.7%, p = 0.0015) compared with the BOV group. CONCLUSIONS: Extremely premature infants who received an exclusive HUM diet had a significantly lower incidence of NEC and mortality. The HUM group also had a reduction in late-onset sepsis, BPD, and ROP. This multicenter study further emphasizes the many benefits of an exclusive HUM diet, and demonstrates multiple improved outcomes after implementation of such a feeding protocol.