Effects of virtual reality training on racket sports performance: A systematic review and meta-analysis of controlled trials

Background: Virtual reality (VR) interventions are increasingly used in racket sports, and quantitative evidence is emerging. Purpose: To synthesise controlled trials examining the effects of VR-based training on racket sports performance outcomes. Methods: A systematic search of five databases (PubMed, Web of Science, Scopus, SPORTDiscus, PsycINFO) up to 15 August 2025 was conducted following PRISMA 2020 guidelines. Eligible studies were controlled trials (randomised or non-randomised) comparing VR-based training (immersive or exergaming) with non-VR controls in tennis or table tennis players. Risk of bias was assessed using Cochrane RoB 2 (randomised) and ROBINS-I (non-randomised). Standardised mean differences (Hedges’ g) were pooled using a random-effects model with Paule–Mandel τ² and Hartung–Knapp–Sidik–Jonkman (HKSJ) adjustment. The protocol was registered (PROSPERO CRD420251132325). Results: Six controlled trials(total N = 426; analysed N = 401) were included. The pooled meta-analysis indicated a moderate overall effect favouring VR (Hedges’ g = 0.78; HKSJ 95% CI [0.41, 1.15]), with moderate heterogeneity (I² = 52%). The 95% prediction interval was [−0.01, 1.57], which spans the null effect, indicating that in some future populations or settings VR training may not yield a meaningful performance advantage. One trial investigated perceptual-cognitive VR training (Anguera et al., 2025), showing a substantial but imprecise effect (g = 0.81, 95% CI [−0.05, 1.67]). As only a single study was available, this evidence is presented descriptively rather than meta-analytically. A subgroup examination of five physically engaging VR interventions found a similarly large effect (*g* = 0.78, 95% CI [0.31, 1.25]) despite considerable heterogeneity (I² = 62%). Conclusion: According to six controlled trials, VR training was associated with average performance enhancements in racket sports; however, the 95% prediction interval [−0.01, 1.57] encompasses the null effect, indicating that a future study could plausibly show no benefit. Due to the limited evidence base (k = 6), significant heterogeneity (I² = 52%), and considerable statistical uncertainty, the existing evidence does not allow for definitive conclusions regarding the efficacy of VR. The current findings should be regarded as a preliminary signal rather than confirmation of effectiveness. Larger, more methodologically robust RCTs with standardised outcomes are needed before definitive recommendations can be made.