BACKGROUND: Preventing anxiety and depression among college students is a pressing public health need. Recent meta-analyses have examined mobile mindfulness interventions in adult populations; however, college students are in a unique developmental stage and institutional setting. METHODS: We conducted a systematic literature review and meta-analysis of published and unpublished studies in English language on the acceptability, usage, and efficacy of mindfulness training apps on mental health among non-clinical samples of college students. Out of 167 reviewed studies, 47 were included in the narrative review. Additionally, we summarized effects from 19 stress, 12 anxiety, 13 depression, and 8 emotional well-being trials (total N = 2974) using robust variance estimation meta-regression and evaluated certainty of evidence with the GRADE approach. RESULTS: Apps were acceptable, with usage levels varying. They reduced stress by 0.435 standard deviation units, 95 % CI (-0.615,-0.255), and increased emotional well-being by 0.431 (0.162,0.7) approaching medium effect sizes. The apps had small effects on depression (B = -0.219 (-0.374, -0.065)) and anxiety (B = -0.218 (-0.42, -0.016)). Certainty of evidence was moderate for stress, depression, and well-being; and low-to-moderate for anxiety. Distressed participants had larger improvements in all outcomes except depression. LIMITATIONS: Small sample sizes in the original studies and small numbers of studies limit the precision of our effect estimates. The small number of studies with objective usage data impedes our ability to characterize the optimal dose. CONCLUSIONS: With moderate certainty of evidence, mindfulness training apps may improve student mental health with similar or larger effect sizes than in the general adult population. However, sustained usage may be a challenge, and more research is needed on the optimal implementation strategy, dose, and equity.