OBJECTIVES: To determine the comparative effectiveness of different interventions for improving antiretroviral medication adherence in low-income settings.  METHODS: We obtained individual patient data from all randomized trials that have evaluated an adherence intervention to promote antiretroviral adherence within low-income countries. We created a treatment network of the differing interventions by pooling the individual patient data from comparable treatments and comparing them across the individual interventions using a Bayesian network meta-analysis approach. Outcomes included self-reported adherence and viral suppression. RESULTS: We obtained data on 11 randomized, involving 5432 patients. Interventions included daily and weekly text messaging, calendars, peer supporters, alarms, counseling, and basic clinical care. For self-reported adherence, we found compelling evidence for the role of weekly text messages (Odds ratio [OR] 1.57, 95% Confidence Intervals [CI] 1.22-2.02), counseling (OR 1.43, 95% CI, 1.06-1.94), and peer supporters (OR 1.72, 95% CI, 1.28-2.29). We found no compelling evidence for daily text messaging, alarms, calendars, or unsupported clinical care. Results were similar when using viral suppression as an outcome, although not all trials reported viral outcomes. Treatment supporters (OR 1.36, 95% CI, 1.02-1.82) and weekly text messages (OR, 1.56, 95% CI, 1.01-2.39) were superior to basic clinical care. CONCLUSIONS: Using individual patient data allowed us to increase precision to determine what interventions appear to work. Several common recommendations for improving adherence are unsupported by the available evidence. These findings should influence guidance documents on improving antiretroviral adherence in poor settings.