BACKGROUND: Treatment of young HIV-infected children is challenging because of rapid disease progression, high viral loads and few drug options. This review was undertaken to update evidence on the management of young HIV-infected children and to inform the development of the 2013 WHO guidelines for antiretroviral therapy (ART) in low and middle-income countries. DESIGN: A systematic review and meta-analysis. METHODS: We identified and critically assessed randomized controlled trials that evaluated treatment strategies in perinatally HIV-infected infants and young children (aged <3 years). RESULTS: Eight studies were included. Antiretroviral therapy (ART) initiation in asymptomatic infants led to 74% reduction in mortality or disease progression [hazard ratio 0.36, 95% confidence interval (CI) 0.18-0.74, P = 0.0002]. Regardless of previous exposure to prevention of mother to child transmission (PMTCT), treatment failure at 24 weeks was more likely in children starting nevirapine-based than in those starting lopinavir/ritonavir (lopinavir/r)-based ART (hazard ratio 1.79, 95% CI 1.33-2.41, P = 0.0001). Infants starting lopinavir/r-based ART and substituting lopinavir/r with nevirapine once virologic suppression was achieved were less likely to experience viral load more than 50 copies/ml (hazard ratio 0.62, 95% CI 0.41-0.92, P = 0.02) but more likely to have confirmed virologic failure (>1000 copies/ml) than those remaining on lopinavir/r (hazard ratio 10.19, 95% CI 2.36-43.94, P = 0.002). Children receiving induction-maintenance ART (four-drug NNRTI-based regimen for 36 weeks followed by three-drug ART) showed better short-term immunologic and virologic responses, but no long-term benefits. The only trial comparing continuous ART from infancy with interrupted ART beyond infancy was terminated early because the duration of treatment interruption was less than 3 months in most infants. CONCLUSION: ART initiation in asymptomatic infants reduces morbidity and mortality. Lopinavir/r-based first-line ART is superior to nevirapine-based regimens in young children, regardless of PMTCT exposure, but lopinavir/r use is challenging. Substituting lopinavir/r with nevirapine following virologic suppression may be feasible where viral load testing is available. Considering current evidence, induction-maintenance and treatment interruption strategies are not recommended. This review contributed to the evidence base for the 2013 WHO guidelines on antiretroviral therapy, which recommend that all children below 3 years start lopinavir/r-based ART and that lopinavir/r can be substituted with nevirapine once sustained virologic suppression is achieved.