Background: Mobile technology gives researchers unimagined opportunities to design new interventions to increase physical activity. Unfortunately, it is still unclear which elements are useful to initiate and maintain behavior change. Objective: In this meta-analysis, we investigated randomized controlled trials of physical activity interventions that were delivered via mobile phone. We analyzed which elements contributed to intervention success. Methods: After searching four databases and science networks for eligible studies, we entered k=49 studies with N=5764 participants into a random effects meta-analysis, controlling for baseline group differences. We also calculated a meta-regression with categories of intervention techniques for behavior change (education, goals, feedback/prompts, rewards, and social cues) as moderators. Results: We found a small overall effect of Hedges' g=.29, (95% CI: .20 to .37) which reduced to g=.22 after correcting for publication bias. None of the predefined elements (e.g., education) led to more intervention success by themselves, but a combination of goals and rewards showed that using goals and rewards simultaneously increased intervention efficacy by g=.31 compared to using none of the two elements. Conclusion: Overall, mHealth interventions to increase physical activity have a small to moderate effect. However, including functions related to goals and rewards led to greater intervention success. More research is needed to look at single behavior change techniques instead of categories. To achieve this, reporting interventions in trial registrations and articles should be more structured and more thorough.