Anxiety and depression are key contributors to the global burden of disease. However, treatments

for these mental health disorders are not always efficacious and there is evidence of a large

treatment gap: affected individuals are often unable or unwilling to access treatment that could

benefit them. Meanwhile, evidence for the beneficial effects of nature on mental health is steadily

building, and the translation of natural environments to virtual reality forms constitute a method to

overcome this issue of accessibility. The present thesis reports on the development and evaluation

of a virtual reality nature intervention for depression and anxiety, facilitated through four studies. In

Studies 1 and 2, practicing psychotherapists (n=9, Study 1) and university students with symptoms of

depression and/or anxiety (n=7, Study 2) trialled and evaluated this intervention, with their feedback

gathered through semi-structured interviews. Results indicated strong potential for the environment

to be used as an intervention, and suggestions for refinement of the intervention/environment were

made and consequently implemented. In Study 3, the potential for a revised version of the

environment to influence short-term affect was measured in individuals with symptoms of

depression and/or anxiety (n=14). A 10-minute session using the environment was found to

significantly decrease negative affect and increase feelings of relaxation (‘serenity’) across the

sample. In Study 4, participants from Study 3 completed a two-week trial where they endeavoured

to use the VR intervention daily. Adherence to the study protocol was high and the group’s

symptoms of depression and anxiety significantly decreased across the study period, though this

remained significant for only depression when compared to a control period. Nevertheless, this

thesis suggests that VR nature could be used in this population for the management of both short-

term affect and longer-term mental health. Suggestions are made for how such an intervention

might best be constructed/implemented.