BACKGROUND: University students are expressing an increased need for mental health support. Mindfulness-based interventions (MBIs) are being integrated into university stress-reduction programmes globally. We conducted a comprehensive systematic review and meta-analysis of randomised controlled trials (RCTs) assessing MBI effects on university students' mental and physical health. METHODS: We searched nine databases, including grey literature and trial registries. Two independent reviewers extracted data following a prospective public protocol. RESULTS: Fifty-one RCTs were included. In comparison with passive controls, and when measured shortly after intervention completion, MBIs improve distress, anxiety, depression, well-being, rumination, and mindfulness with small to moderate effect sizes, with no benefit found for blood pressure, sleep, life satisfaction, resilience, worry, and thought suppression. Evidence for self-compassion is inconclusive. Effects last beyond three months for distress and mindfulness, with no data on other outcomes. Compared with active control groups, MBIs significantly improve distress and state anxiety, but not mindfulness, depression, well-being, affect, trait anxiety, or emotion regulation. Results were robust to adjustment for multiple testing, but RCTs' risk of bias is generally high. Moderator analyses did not find differential intervention effects according to intervention duration, delivery mode, or sub-populations. CONCLUSIONS: MBIs may be helpful to students but higher-quality research is needed.