Abstract
<title>Abstract</title> <p>Rising mental health needs and limited access to restorative environments have increased interest in scalable, technology-based interventions. Immersive virtual reality (VR) can simulate natural environments, yet evidence on its combined psychological and physiological effects remains fragmented. This systematic review synthesises empirical studies on the impact of immersive virtual nature (IVN) on mental health and stress-related biomarkers in adults. A systematic search of PubMed, Scopus, Web of Science, and APA PsycINFO databases identified 12 peer-reviewed studies. Inclusion criteria required the use of head-mounted displays (HMDs) simulating natural environments and the reporting of both psychological measures and physiological biomarkers in an adult population. Risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB 2) tool for randomised studies and the Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I) tool for non-randomised studies. Results showed that IVN reduced psychological stress and negative affect, with medium effect sizes (d = 0.26–0.64). Improvements in cardiovascular regulation and neurological markers of relaxation were observed in 80–100% of relevant studies. Greater benefits were found in clinical and occupational populations, older adults, and individuals with anxiety. Multisensory environments and nature scenes including water or semi-open green landscapes showed better outcomes, while repeated identical exposures had diminished effects over time. The findings support the use of IVN as a scalable intervention for short-term stress relief and emotional regulation, especially where real nature is inaccessible. Limitations include the short duration of most interventions, variability in outcome measures, and a lack of studies in underrepresented or high-need populations. Future research should prioritise longitudinal designs, standardised protocols, and focus on clinical populations. This review was retrospectively registered on the Open Science Framework (OSF: osf.io/hj3pc) prior to screening, data extraction, and synthesis.</p>