BACKGROUND: Antiretroviral therapy is effective in preventing the progression of HIV to AIDS, but adherence to HIV medication is lower than ideal. A previous Cochrane review concluded that SMS interventions increased adherence to HIV medication, but more recent trials have reported mixed results. Our review aims to provide an up-to-date synthesis of the effects of interventions delivered by mobile phone on adherence. METHODS: We searched Cochrane, Medline, CINAHL, EMBASE and Global Health for randomised control trials (RCTs) of interventions delivered by mobile phones, designed to increase adherence to antiretroviral medication. Risk of bias was assessed using the Cochrane risk of bias tool. We calculated relative risk ratios (RR) or standardised mean difference (SMD) with 95% confidence interval (CI). Trials were analysed depending on delivery mechanism and intervention characteristics. We conducted meta-analysis for primary objective outcome measures. RESULTS: We identified 19 trials. No trials were at low risk of bias. Interventions were delivered as follows; nine via text message, five via mobile phone call, one via mobile phone imagery and four via mixed interventions. There was no effect when interventions delivered by text message were pooled in the RR1.25 (CI 0.97 to 1.61) P = 0.08. The SMD 0.42 (0.03 to 0.81) p = 0.04 showed a moderate effect to improve adherence. There was mixed evidence of the effect of text messages delivered daily, weekly, at scheduled or triggered times, however, messages with link to support, interactivity and three or more behavior change techniques (BCTs) all improved adherence. Of the five trials delivered by mobile phone call, one reported a reduction in HIV viral load. One trial using mobile phone imagery reported a reduction in HIV viral load. Three trials that delivered interventions by text message and mobile phone counselling reported improved biological outcomes. CONCLUSION: Specific interventions, of proven effectiveness should be considered for implementation, rather than mobile phone-based interventions in general. Interventions targeting a wider range of barriers to adherence may be more effective than existing interventions. The effects and cost-effectiveness of such interventions should be evaluated in a randomised controlled trial alongside long term objective and clinically important outcomes.