Virtual reality nature exposure reduces self-criticism and shows trends toward enhanced self-compassion: A randomized...

Building on the therapeutic potential of nature exposure and the increasing accessibility of immersive virtual reality (VR), this study investigates whether custom-designed VR environments-specifically a virtual forest and a virtual city-can differentially influence self-compassion, self-criticism, self-protection, and perceived stress.

This randomized controlled trial aimed to evaluate the psychological impact of VR nature exposure compared to VR urban exposure on self-related emotional processes in a non-clinical adult sample. This study extends our previous pilot work (Szitás et al., 2025) with a larger sample, refined virtual environments, and enhanced methodological rigor.

In a between-subjects design, 34 participants (mean age = 30.25 years, SD = 7.5; 55.9% women) were randomly assigned to either a virtual forest group (n = 17) or virtual city group (n = 17). Participants completed two 25-min VR exposure sessions (preceded by a 2-min acclimatization phase) within seven days using a Meta Quest 2 head-mounted display in a controlled laboratory setting. Self-report measures of state self-compassion, self-criticism, self-protection, and perceived stress were collected at baseline, post-intervention, and follow-up. A sensitivity power analysis indicated that with N = 34 and α = 0.05, the study had 80% power to detect medium-to-large effects (Kendall's W ≥ 0.25). Data were analyzed using Friedman tests with Durbin-Conover pairwise comparisons.

The forest group showed a significant reduction in state self-criticism across time points (χ²(4) = 21.6, p < .001, Kendall's W = 0.32), with scores decreasing from baseline (M = 48.9) to post-first-exposure (M = 35.7). Self-compassion showed a non-significant upward trend (χ²(4) = 8.00, p = .092, W = 0.12). No significant changes occurred in the city group for any measure (all p > .05). Trait-level measures, including perceived stress and self-protection, remained stable in both groups.

Brief VR nature exposure significantly reduced momentary self-criticism and showed a trend toward enhanced self-compassion. These findings support the potential of VR-based nature interventions as accessible digital mental health tools, particularly for individuals with limited access to natural environments.