Studies have confirmed a major reduction in necrotizing enterocolitis (NEC) and death with routine use of probiotics in high-risk preterm infants. Although no major differences were reported in the preparations containing lactobacilli, just bifidobacteria, and a mixture, mixtures of different organisms may be more effective than a single species. This prospective cohort study was performed to determine the clinical outcomes after starting probiotics use in a neonatal intensive care unit (NICU). After the first 17 months of using probiotics, incidences of NEC and death were compared with those during the prior 17 months before introducing probiotics. All infants at less than 32 weeks’ gestation at birth were eligible to receive probiotics unless they had a gastrointestinal malformation. They received FloraBABY probiotic mixture, containing Bifidobacterium breve, Bifidobacterium bifidum, Bifidobacterium infantis, Bifidobacterium longum, and Lactobacillus rhamnosus GG (2 × 109 colony-forming units/0.5 g) mixed with 1 mL water just before milk feeding once daily. Probiotics were started at the time of the first feed and continued until the infant reached 34 weeks’ postmenstrual age. Mortality was defined as death before discharge home. Health care–associated infection (HCAI) was defined by a positive culture of a normally sterile body fluid, occurring after 3 days of age. As a proxy for feeding tolerance, the day that intravenous (IV) nutrition was first stopped and the final day of receiving IV nutrition were recorded. The preprobiotic and probiotic cohorts included 317 and 294 infants, respectively. Compared with the preprobiotic cohort, the probiotic cohort had a significant reduction in the incidence of NEC (10% vs 5%; P < 0.05), a nonsignificant reduction in death (10% vs 7%), and a significant reduction in the proportion of infants who died or developed NEC (17% vs 11%; P < 0.05). No significant effect on HCAI (18% vs 17%) was noted. The average duration of therapy was 24 ± 16 days, and the average time of starting probiotics was day 4.3 ± 5.1. When only the infants with a birth weight of less than 1001 g were analyzed, similar percentage reductions in NEC and the combined outcome of death and NEC were found. The 2 cohorts did not differ significantly in average gestational age or in proportions who were small for gestational age, female, or less than 26 weeks’ gestation. The odds ratios for NEC, death, NEC or death, and HCAI were 0.51 (95% confidence interval [CI], 0.26–0.0.98), 0.71 (95% CI, 0.38–1.34), 0.56 (95% CI, 0.33–0.93), and 0.98 (95% CI, 0.81–1.98), respectively, with receipt of probiotics. The mean difference in the time to first stopping IV nutrition was 4.6 days (95% CI, 1.8–7.3 days), and the mean difference in time to finally stopping IV nutrition was 5.3 days (95% CI, 0.8–9.9 days). Administration of probiotic organisms to prevent NEC is feasible, supported by clinical evidence, and has a beneficial balance. Studies comparing strains and examining additional NICU populations and length of therapy are needed.