Does Virtual Reality Therapy Work? The Evidence for VR in Mental Health
Virtual reality therapy sounds like the sort of thing a tech company would describe as “the future of mental health” while quietly hoping nobody asks about sample sizes.
That does not mean it is nonsense.
VR therapy is one of the more interesting developments in psychological treatment, especially for problems where people need to practise facing feared, avoided, painful, or overwhelming situations in a controlled way. It can create environments that are difficult, expensive, impractical, or unsafe to arrange in real life: flying, heights, crowds, driving, public speaking, hospital procedures, social situations, or everyday spaces that have become frightening.
The trouble is that VR attracts two equally unhelpful reactions.
One side treats it as magic goggles. Put on the headset, enter the future, emerge healed and softly lit. Lovely, if you are selling a conference keynote.
The other side dismisses it as a gimmick because it looks technological, immersive, and faintly ridiculous to anyone who remembers early VR demos where everyone looked like they were being attacked by a screensaver.
The evidence sits somewhere more useful. VR therapy is not a revolution in all of psychiatry. It is a delivery method. A tool. Sometimes a very clever one. It appears most useful when it strengthens an existing therapeutic principle, especially exposure therapy, rather than pretending the headset itself is the treatment.
That distinction saves us from a lot of nonsense.
What is virtual reality therapy?
Virtual reality therapy uses immersive digital environments as part of psychological or medical treatment.
A person wears a headset or uses another immersive display and enters a simulated environment. Depending on the treatment, that environment might be a plane cabin, a busy shopping centre, a tall building, a social gathering, a hospital room, a street, a classroom, or a calm landscape designed for pain or anxiety management.
The point is not simply to impress the patient with graphics. If therapy only required impressive graphics, most video games would be clinically regulated and teenagers would be paragons of emotional balance. This has not occurred.
The real value of VR is that it can make therapeutic experiences controllable, repeatable, gradual, and safe enough to practise.
A therapist can increase or reduce the intensity of the situation. A patient can repeat the same task several times. Exposure can begin gently and build up. Difficult real-world situations can be simulated without the cost, risk, logistics, or public embarrassment of arranging the real thing too early.
For some conditions, this is genuinely useful.
For others, the evidence is thinner, the applications are more speculative, and the enthusiasm has sprinted ahead of the data like it heard there was venture funding nearby.
Where the evidence is strongest: exposure therapy for anxiety and phobias
The strongest case for VR in mental health is exposure-based treatment.
Exposure therapy helps people gradually face feared situations, sensations, memories, or objects in a safe and structured way. The aim is not to terrify the person for character-building reasons, because therapy is not a boarding school in 1920. The aim is to help the nervous system learn that the feared situation can be tolerated, managed, and reinterpreted.
VR fits exposure therapy very naturally.
Someone with a fear of heights can stand on a virtual balcony. Someone afraid of flying can sit in a simulated plane. Someone with public speaking anxiety can practise in front of a virtual audience. Someone with social anxiety can rehearse social encounters in a controlled environment before trying them in real life.
This is where VR is at its most sensible. It does not replace the theory of treatment. It makes certain forms of exposure easier to deliver.
Recent reviews continue to suggest that VR exposure therapy is especially promising for specific phobias and social anxiety. One 2025 systematic review of randomised controlled trials found VR treatment particularly effective for specific phobias and social anxiety disorder, while outcomes for generalised anxiety disorder, agoraphobia, and panic disorder were less robust. Therapist involvement also appeared to improve effectiveness.
That is a more useful conclusion than “VR treats anxiety.”
Anxiety is not one thing. Fear of spiders, public speaking anxiety, panic disorder, agoraphobic avoidance, trauma-related hyperarousal, and generalised worry are different problems. VR may help some of them more than others.
Why VR exposure can be useful
VR exposure has several practical advantages.
It gives clinicians control. A therapist can decide whether the virtual plane is boarding, taking off, cruising, or hitting turbulence. A virtual audience can be quiet, mildly distracted, or openly unimpressed in a way that feels cruelly realistic but remains clinically adjustable.
It improves access to difficult situations. You cannot easily bring a plane, motorway, crowded tube station, or high-rise balcony into a therapy room. Well, not without a budget, planning permission, and a nervous receptionist.
It allows repetition. The same feared situation can be practised again and again, which is central to learning.
It can feel emotionally real enough to trigger useful therapeutic work, while still being safe enough for the person to stay engaged.
It may also reduce the barrier to beginning exposure. Some people are not ready to go straight into a real feared situation. VR can create an intermediate step between imagination and real-world exposure.
This does not mean VR is always better than traditional exposure. Sometimes real-world exposure is cheaper, more direct, and more generalisable. But VR can help where real-world exposure is too difficult to arrange, too unpredictable, or too overwhelming at the start.
PTSD and trauma-related work
VR has been studied in PTSD and trauma-related presentations, often as a way to support controlled exposure to trauma-related cues.
This is an area where caution is needed.
On one hand, VR can create structured environments that help people approach avoided memories or situations with clinical support. That may be useful for some trauma-related problems, particularly when avoidance is maintaining distress.
On the other hand, trauma work is not improved simply by making it more immersive. A headset is not a shortcut through the careful work of stabilisation, consent, pacing, safety, meaning, and therapeutic relationship.
There is a familiar technological temptation here: if something is difficult in therapy, make it more vivid. That can be useful. It can also be a superb way to overwhelm someone if handled badly.
VR for PTSD should be used by properly trained clinicians, with careful screening, clear rationale, and appropriate safeguards. It should not be sold as trauma treatment because the graphics are impressive and someone has discovered the word “immersive.”
Psychosis and agoraphobic avoidance: the gameChange example
One of the most interesting modern developments is the use of automated VR therapy for people with psychosis who experience agoraphobic avoidance and distress in everyday situations.
The gameChange trial, published in The Lancet Psychiatry in 2022, tested an automated VR cognitive therapy across nine NHS trusts in England. The therapy was designed to help people with psychosis practise everyday situations that felt frightening or avoided, such as entering a shop, using public transport, or being around other people.
This is important because it shows VR being used for something more clinically specific than generic “mental health innovation.”
The target was not psychosis as a whole. It was agoraphobic avoidance and distress in people with psychosis. That distinction keeps the claim honest.
Oxford’s summary of the project reports that gameChange led to significant reductions in avoidance of everyday situations and distress, with the strongest benefits among patients with severe agoraphobia and more severe psychiatric symptoms, and benefits maintained at six-month follow-up.
That is a good example of where VR may be genuinely useful: helping people practise feared real-world situations in a graded, repeatable way, while reducing the burden on therapist time through partial automation.
It is not “VR cures psychosis.” It is more specific and more interesting than that.
Pain and medical settings
VR is also used outside psychiatry, especially in pain and medical settings.
The logic is fairly clear. Pain is shaped by attention, fear, emotion, expectation, and context. VR can absorb attention, reduce threat, provide distraction, and help people feel less trapped inside the pain experience.
VR has been studied in burn care, paediatric procedures, chronic pain, dental procedures, and other medical contexts. A 2024 review article described VR as a complementary or adjunctive intervention for pain and anxiety, particularly in paediatric and medical settings.
That does not mean VR is a replacement for pain medication, anaesthesia, physiotherapy, psychological pain treatment, or proper medical care. It means VR may be useful as part of a broader approach.
The same rule applies here as with hypnosis: if someone says “this tool can help some people manage pain,” listen. If someone says “this tool removes pain and makes medicine unnecessary,” leave the room before they find a brochure.
Autism, ADHD, and neurodevelopmental uses
The old version of this article said VR shows promise for “treating” ADHD and autism.
That wording needs softening.
VR is being explored for neurodevelopmental contexts, including social skills practice, attention training, assessment, executive function tasks, emotional recognition, and controlled rehearsal of real-world scenarios. That is interesting. It may be useful. But calling it a treatment for ADHD or autism is too broad.
Autism is not a disease to be cured by a headset. ADHD is not a simple lack of attention that can be fixed by making tasks more visually exciting. Both involve complex neurodevelopmental profiles, social contexts, individual needs, strengths, difficulties, and support requirements.
A better way to frame this is: VR may become useful in specific educational, clinical, or skills-practice contexts, especially where realistic but controlled environments are valuable. But the evidence base is not as strong as it is for exposure-based treatment of phobias and social anxiety.
The hype here needs a shorter lead.
Depression and broader psychiatric treatment
VR has been studied in depression, but this is not the strongest part of the evidence base.
Some VR interventions aim to increase behavioural activation, provide immersive relaxation, practise self-compassion, simulate social interaction, or support engagement in therapy. These are plausible ideas. They may help some people. But depression is highly varied, and VR is not obviously central to treating it in the way it is naturally suited to exposure-based anxiety work.
The 2020 systematic review of reviews by Cieślik and colleagues concluded that VR therapy showed a positive impact across psychiatric disorders, but also noted that the impact differed depending on the clinical area and that further research was needed.
That should be the tone here. Interesting, not triumphant.
VR may be a useful addition in some depression treatments, especially where engagement, social practice, behavioural activation, or embodied simulation are relevant. But it should not be presented as the next great replacement for established therapies, medication, social support, or the less fashionable intervention of making people’s lives less miserable.
Why VR might help
VR therapy may help through several mechanisms.
The first is presence. Presence is the feeling of being “there” in the virtual environment. If the brain responds to the simulation as meaningful, the person can practise emotional and behavioural responses in a way that feels more real than imagination alone.
The second is controlled exposure. VR allows therapists to control intensity, repeat scenarios, and build up gradually.
The third is learning through practice. People do not just talk about entering a feared situation. They experience something like it and practise responding differently.
The fourth is safety. A person can face a virtual lift, plane, spider, audience, or crowded street without being trapped in the real-world version.
The fifth is accessibility and scalability. Some VR therapies can be automated or partly automated, potentially increasing access where therapist time is limited. The gameChange trial is a useful example of this direction.
None of this means VR is inherently therapeutic. A headset is not a therapist. A virtual environment is not a treatment plan. The clinical design matters. The target problem matters. The therapist or system guiding the work matters.
Otherwise, it is just someone standing in a digital shopping centre feeling awful in higher resolution.
Limits and risks
VR therapy has real limitations.
Some studies are small. Some have weak control groups. Some rely on short follow-ups. Some outcomes are measured immediately after treatment, which is useful but not enough. A person feeling better after an immersive intervention is encouraging. The more important question is whether change lasts and transfers into real life.
There is also a novelty effect. New technology can feel exciting, engaging, and convincing simply because it is new. That does not mean the effect will survive once the headset stops feeling futuristic and starts feeling like another clinical appliance that needs wiping.
VR can also cause cybersickness, nausea, dizziness, eyestrain, or discomfort. Some people dislike headsets. Some find them claustrophobic. Others may struggle with sensory overload.
Access is another issue. Hardware costs, software quality, clinician training, privacy, data security, and maintenance all matter. Mental health services are not usually drowning in spare money, spare time, or spare staff. Adding technology is only useful if it improves care rather than becoming another underfunded gadget cupboard.
Privacy also deserves attention. VR systems can collect sensitive behavioural, movement, emotional, and physiological data. In mental health contexts, that data needs proper protection. A person’s panic response in a virtual supermarket should not become a business opportunity with terms and conditions nobody reads.
Finally, VR can be overused or badly targeted. More immersion is not always better. More exposure is not always helpful. And technology should not become a way to avoid the slower, less flashy work of therapeutic relationship, formulation, social support, and real-world change.
Is VR better than ordinary therapy?
Sometimes, possibly.
Often, not necessarily.
VR may be better than ordinary therapy when the treatment requires repeated exposure to situations that are hard to arrange in real life. It may also improve engagement for some people, especially when traditional exposure feels too abstract, too logistically difficult, or too intimidating.
But VR is not automatically superior to face-to-face therapy, CBT, medication, family work, trauma therapy, group interventions, or real-world exposure.
The 2025 systematic review of RCTs on immersive VR for anxiety disorders found VR treatment particularly effective for specific phobias and social anxiety, with comparable efficacy to non-VR treatments in some areas. That is promising, but “comparable” is not the same as “obviously better.”
This is where the marketing can get slippery.
If VR is as effective as an established treatment but easier to access, cheaper to scale, or more acceptable to some patients, that is still valuable. A treatment does not have to be dramatically superior to be clinically useful. Sometimes being easier to deliver is enough.
But if VR is more expensive, less accessible, poorly supported, or used without proper clinical reasoning, then the headset is just a very modern way to avoid doing the basics well.
Where VR therapy is heading
VR therapy is likely to become more common.
Headsets are improving. Software is becoming more flexible. Automated programmes are being tested. Clinicians are learning how to use virtual environments more thoughtfully. Research is moving beyond simple phobia scenarios into psychosis, social anxiety, pain, rehabilitation, training, and complex clinical presentations.
That is promising.
But the future of VR therapy should not be treated as inevitable. Mental health care has seen plenty of “future of treatment” claims wander in wearing shiny shoes and leave behind a disappointing evidence base.
The best future for VR therapy is not as a replacement for clinicians or a universal mental health platform. It is as a targeted clinical tool used when immersion, rehearsal, exposure, repeatability, and controlled environments genuinely add something.
That is less dramatic than “revolutionising psychiatry.”
It is also more likely to be true.
Simply Put
Virtual reality therapy can work, but not because VR is magic.
It works best when it gives therapy something useful: controlled exposure, realistic practice, repeatable scenarios, and safe rehearsal of difficult situations. That makes it especially promising for specific phobias, social anxiety, and some anxiety-related problems.
It is also being used in more specialised ways, such as automated VR cognitive therapy for agoraphobic avoidance in people with psychosis. The gameChange trial is one of the strongest examples of VR being tested in a serious clinical context rather than just waved around as futuristic decoration.
The evidence is less clear for broader claims about depression, ADHD, autism, and psychiatric treatment in general. VR may have uses in those areas, but it should not be sold as a cure, a replacement for therapy, or a technological shortcut around proper clinical care.
The best way to understand VR therapy is simple: it is a tool for creating experiences that therapy can use.
Sometimes that is powerful.
Sometimes it is overhyped.
And sometimes it is just an expensive headset looking for a problem to justify its existence.
References
Table of Contents
