BACKGROUND: The potential of using mobile phone technologies to improve antiretroviral therapy (ART) adherence has provided a new facet to human immunodeficiency virus (HIV) research. The quality of evidence and the strength of recommendations of existing reviews, however, do not adequately support large-scale adoption of the intervention. This review adopted broad selection criteria to include all mobile phone-based interventions designed to improve patient's adherence to ART. METHODS: We performed a systematic review and meta-analysis of randomized controlled trials and quasi-experimental studies. PUBMED, MEDLINE, EMBASE, PsychINFO, Cochrane Central Register of Controlled Trials (CENTRAL), CINAHL, AMED and Web of Science were searched. Online abstracts archives of relevant conference proceedings and trial registries were also searched. Thirty-Five (35) full-text articles were assessed for eligibility. Included studies were conducted in high, low and middle-income countries and reported ART adherence interventions delivered by mobile phones (standard or smartphones) in the form of voice calls, interactive voice response calls (IVR), and short message service (SMS). RESULTS: Thirteen (13) studies met the inclusion criteria, and 11 were used in the meta-analysis. Intervention characteristics of included studies ranged from mobile phone functionalities to provision of study phones to participants. SMS and voice call contents were tailored to participants' specific adherence needs. Mobile SMS interventions improved adherence to ART compared with control conditions (OR, 95% CI = 1.59, 1.27-1.98). In subgroup analysis, only scheduled SMS was significant whereas triggered SMS had no effect on adherence to ART. Mobile voice calls did not significantly increase adherence to ART. The interventions were highly rated by > 90% of participants in the studies that reported on the experiences and satisfaction with the intervention. CONCLUSION: Scheduled mobile phone text-messaging have demonstrated significant improvement in adherence to ART. Mobile SMS adherence interventions that allow for two-way communication may, however, be more acceptable than standalone SMS reminders, which are seen to be intrusive, producing habituation and response fatigue. Voice calls and triggered SMS functionalities do not have a significant effect on adherence to ART although there is a higher preference for voice functionality over SMS especially in limited-resource and low-literacy settings. Further exploration of the mobile voice functionality and its possible combination with scheduled SMS functionality is recommended. Evidence provided in this study will guide the implementation of mobile phone intervention to improve adherence to ART, by addressing practical challenges that could militate against scalability especially in resource limited settings.