We previously conducted two randomized controlled trials with bovine lactoferrin (bLF) for the prevention of lateonset sepsis (LOS) in infants with a birth weight<2500 g (Study 1) and<2000 g (Study 2). The aim of this study was to determine the preventative effects of bLF on culture-proven or probable LOS in infants with a birth weight<1500 g from both studies, and to determine the effect of bLF in relation to intake of human milk. Both trial designs had similar inclusion and exclusion criteria, the same dose of bLF [200 mg·(kg body mass)-1·day-1], and used the same control (maltodextrin). We fitted multivariate Cox regression models to estimate the effect of bLF on the risk of development of the composite outcome, adjusting for covariates. We included 335 neonates with a mean birth weight of 1162 6 244 g; 27.5% were<1000 g. There were 33 first episodes of LOS in the bLF treatment group and 48 in the control group (19.5% vs. 28.9%). bLF had a protective effect on the risk of development of LOS [hazard ratio (HR)=0.64; %95 CI=0.41-0.99; p=0.048]; particularly among infants weighing<1000 g [HR=0.46; %95 CI=0.22-0.96; p=0.039] and infants with a low intake of human milk [HR=0.40; %95 CI=0.19-0.84; p=0.015]. Therefore, bLF supplementation protects infants<1500 g from LOS, particularly those infants not receiving human milk.