Background: Perinatal transmission of HIV has dramatically decreased in high-income countries in the last few years with current rates below 1%, but it still occurs in high-risk situations, mainly pregnant women with late diagnosis of infection, poor antiretroviral adherence and a high viral load (VL). In these high-risk situations, many providers recommend combined neonatal prophylaxis (CNP). Our aim was to evaluate the safety and toxicity of CNP in infants deemed at high-risk of HIV infection among motherinfant pairs in the Madrid Cohort. Materials and Methods: Prospective, multicenter, observational cohort study between years 2000 and 2019. The subgroup of newborns on CNP and their mothers were retrospectively selected (cohort A) and compared with those who received monotherapy with zidovudine (cohort B). Infants with monotherapy were classified according to treatment regimes in long (6 weeks) and short (4 weeks) course. Results: We identified 227 newborns (33.3% preterm and 7 sets of twins) with CNP. A maternal diagnosis of HIV-1 infection was established during the current pregnancy in 72 cases (36.4%) and intrapartum or postpartum in 31 cases (15.7%). Most infants received triple combination antiretroviral therapy (65.6%; n = 149). The perinatal transmission rate in cohort A was 3.5% (95% confidence interval: 1.13%-5.92%). Infants from cohort A developed anemia (26.1% vs. 19.4%, P = 0.14) and neutropenia more frequently at 50-120 days (21.4% vs. 10.9%, P < 0.01), without significant differences in grade 3 and 4 anemia or neutropenia between the two cohorts. There were no differences in increased alanine aminotransferase. Neutropenia was more common in the long zidovudine regimes. Conclusions: Our findings provide further evidence of the safety of CNP in infants with high-risk of HIV-1 perinatal transmission.