Objective: The aim of this study was to evaluate the effectiveness of virtual reality interventions for improving balance and gait in people with Parkinson's disease. Design: This is a systematic review and meta-analysis of randomized controlled trials. Methods: Databases of MEDLINE, Cochran Central Register of Controlled Trials, EMBASE, PEDro, Web of Science and China Biology Medicine disc were searched from their inception up to 1 March 2019. Two reviewers individually appraised literatures for inclusion, extracted data and evaluated trial quality. Results: A total of 12 studies with a median PEDro score of 6.4 and involving 419 participants were included. This review first demonstrated significant improvements in Berg Balance Scale (mean difference = 2.69; 95% confidence interval = 1.37 to 4.02; p < 0.0001), Timed Up and Go Test (mean difference = -2.86; 95% confidence interval = -5.60 to -0.12; p = 0.04) and stride length (mean difference = 9.65; 95% confidence interval = 4.31 to 14.98; p = 0.0004) in Parkinson patients who received virtual reality compared with controls. However, there was no significant difference in gait velocity and walk distance. Conclusion: This systematic review and meta-analysis supports the use of virtual reality to enhance the balance of patients with Parkinson's disease. However, the review does not find any definite effect upon gait by the use of virtual reality.