Xianyou Cui, Junhua Tang, Yuxuan Ying, Meng Tao, Hong Sun, Hailong Jiao
Objective This study is aimed at the assessment of the efficacy of virtual reality (VR) technology in improving balance ability among patients with stroke and analyze the potential moderating effects of participant characteristics, training programs, and other factors on interventions. Methods An exhaustive search for published papers up to February 2025 was conducted across five electronic databases to identify randomized controlled trials (RCTs) investigating the effects of VR on balance ability in stroke patients. A systematic review and meta-analysis were performed to synthesize effect sizes using the standardized mean difference (SMD, Hedges’g). Subgroup analyses were conducted to explore sources of heterogeneity. Results Meta analyses identified four principal findings. (1) BBS: VR significantly improved scores across 14 studies (SMD = 0.55, 95% CI: 0.23–0.88, P < 0.05), with the VR+CT subgroup showing the largest effect (SMD = 0.89). (2) TUG: VR significantly enhanced performance in 17 studies (SMD = −0.29, 95% CI: −0.53 to −0.05, P < 0.05); the VR+CT subgroup was optimal (SMD = −0.66, I 2 = 36%). Egger’s test indicated publication bias (P < 0.05), and trim-and-fill correction reduced the effect size. (3) FRT: VR showed no significant benefit in 4 studies (SMD = 0.36, 95% CI: −0.06 to 0.77, P > 0.05). (4) 10MWT: VR showed a borderline, non-significant positive trend in 6 studies (SMD = 0.37, 95% CI: 0.00–0.74, P = 0.05, I 2 = 36%). Conclusion VR intervention may improve selected balance-related outcomes in stroke patients, especially when combined with conventional rehabilitation. VR significantly enhances BBS and TUG outcomes, shows a borderline positive trend on 10MWT, but has no significant effect on FRT. With good safety and high compliance, VR is a feasible adjunct for stroke rehabilitation. The evidence certainty is moderate to low; future large−sample, high−quality RCTs are needed to verify long−term efficacy and support clinical application. Systematic review registration https://www.crd.york.ac.uk/prospero/ , identifier CRD420251266836.