Virtual reality (VR) may enhance the effectiveness of psychological interventions for acute pain. We conducted a meta-analysis to assess the efficacy and safety of VR-based interventions for pain associated with medical procedures.We searched PubMed, EMBASE, the Cochrane Library, and PsycINFO until June 17th 2018. We identified randomized controlled trials (RCTs), comparing VR-based psychological interventions to usual care, for pain intensity (primary outcome) or affective and cognitive components of pain (secondary outcomes), assessed real-time or retrospectively. Two independent reviewers performed study selection and data extraction. Risk of bias was independently evaluated by three raters using the revised Cochrane Collaboration tool. A random-effects model using the Paule and Mandel estimator was used for pooling effect sizes.27 RCTs (1452 patients) provided enough data for meta-analysis. Compared to usual care, VR-based interventions reduced pain intensity both real-time (9 RCTs, Hedges' g = 0.95, 95% CI 0.32-1.57) and retrospectively (22 RCTs, g = 0.87, 95% CI 0.54-1.21). Results were similar for cognitive (8 RCTs, g = 0.82, 95% CI 0.39-1.26) and affective pain components (14 RCTs, g = 0.55, 95% CI 0.34-0.77). There was marked heterogeneity, which remained similarly high in sensitivity analyses. Across domains, few trials were rated as low risk of bias and there was evidence of publication bias. Adverse events were rare.Though VR-based interventions reduced pain for patients undergoing medical procedures, inferring clinical effectiveness is precluded by the predominance of small trials, with substantial risk of bias, and by incomplete reporting.