Virtual Reality in Vision Therapy: Promise and Evidence

Headsets have arrived in vision therapy rooms, ours included. Parents ask about them constantly, usually after seeing a claim online that virtual reality has “revolutionized” treatment for a lazy eye. It is a fair question, and it deserves a straight answer rather than a sales pitch.

Here is ours. Virtual reality gives us a genuinely new way to present a task to two eyes at once, and there is real research behind it. No professional body recommends it as a replacement for glasses or patching — although the American Academy of Ophthalmology now lists digital therapeutics among the suitable treatment options for amblyopia (AAO Amblyopia Preferred Practice Pattern, 2023). This is what the evidence supports, where it stops, and how we use technology in our Coconut Creek, Boynton Beach and West Palm Beach offices.

Key takeaways

• VR headsets do one thing paper cannot: show each eye a different image and adjust the balance between them in real time.
• The adult evidence is early. A 2017 pilot of 17 adults reported better acuity and depth perception, but had no control group (Žiak et al., 2017).
• The first pivotal children’s trial compared a dichoptic therapy with glasses alone, not with patching (Xiao et al., 2022).
• Head-to-head against patching, the results are mixed. A binocular iPad game trailed patching in two large trials (Holmes et al., 2016; Manh et al., 2018) and beat nothing at all in a placebo-controlled trial (Gao et al., 2018) — but a newer eye-tracking home system was noninferior to two hours of daily patching (Wygnanski-Jaffe et al., 2023).
• Adherence is a major variable. A 2026 meta-analysis found dichoptic therapy no better than patching overall, and better in children who actually used it — evidence the authors themselves graded very low certainty (Carvalho Soares Valentim et al., 2026).

What a headset actually does

The idea behind most VR vision therapy is dichoptic presentation: each eye sees its own image at the same moment. Software can then dim or reduce the contrast of what the stronger eye sees, so that the weaker eye has to contribute for the game to make sense. Rebalance the contrast as the patient improves, and you have a task that trains the two eyes as a team instead of covering one and training the other alone.

That distinction matters clinically. In amblyopia, the deficit people notice most in ordinary life is depth perception, and laboratory work suggests that patients with strabismic amblyopia rarely improve with monocular training but do better with dichoptic and stereo-based approaches (Levi et al., 2015). A headset also solves an old, unglamorous problem: children finish a game and often will not finish a worksheet.

Engagement is not the same as effectiveness, though. That is where the trials come in. For the longer story of how the field arrived here, see the history of vision therapy.

What the research shows

Adults: promising, and preliminary

The most cited VR study gave 17 adults with anisometropic amblyopia eight 40-minute sessions in an Oculus Rift headset. Acuity in the amblyopic eye improved from logMAR 0.58 to 0.43, and stereoacuity from 263.3 to 176.7 arcseconds (Žiak et al., 2017). That is an encouraging result in a group long assumed to be past treatment age — but it was an uncontrolled pilot with no comparison group, so practice effects and repeat testing cannot be ruled out. The authors said as much and called for proper trials.

Children: one clear positive, with an asterisk

The first pivotal dichoptic trial in children randomized 105 four-to-seven-year-olds at 21 US sites. A dichoptic digital therapeutic used an hour a day, six days a week alongside glasses improved amblyopic-eye acuity 1.8 lines at 12 weeks, versus 0.8 lines with glasses alone (Xiao et al., 2022). That is a genuine randomized result, and the trial was stopped early for success. The asterisk is the comparator: it was tested against glasses alone, not against patching, so it does not show that a headset beats the standard treatment. The trial was also industry-sponsored, with several authors employed by the manufacturer.

Head-to-head against patching: mixed

Two large trials by the Pediatric Eye Disease Investigator Group put a binocular iPad game directly against part-time patching. In 385 children aged 5 to 12, the game improved acuity 1.05 lines versus 1.35 lines for two hours of daily patching; the trial’s planned noninferiority analysis was formally indeterminate, and a later post hoc analysis suggested the game did worse than patching (Holmes et al., 2016). In 100 teenagers, it was 3.5 letters versus 6.5, a difference that did not reach statistical significance (Manh et al., 2018). And in the placebo-controlled BRAVO trial, an active binocular game and a placebo game produced essentially the same change (Gao et al., 2018).

Newer hardware has done better. A masked randomized noninferiority trial of 103 children aged 4 to under 9 compared an eye-tracking home dichoptic system, used 90 minutes a day five days a week, with two hours of daily patching. Acuity improved 0.28 logMAR with the device versus 0.23 logMAR with patching, meeting the trial’s noninferiority criterion (Wygnanski-Jaffe et al., 2023). So “digital therapy always loses to patching” is not what the literature says. What it says is that the answer depends on the system, the age group and how much of the treatment actually gets done.

Professional guidance has moved with that evidence. In 2020 the American Academy of Ophthalmology reviewed 20 studies and found no level I evidence to support binocular treatment as a substitute for patching or optical correction (Pineles et al., 2020); that search closed in April 2019, before the newer trials. The Academy’s current amblyopia guideline now lists digital therapeutics among the suitable treatment options alongside optical correction, patching, atropine and Bangerter filters, while still concluding that the evidence remains limited to recommend binocular therapy (AAO Amblyopia Preferred Practice Pattern, 2023). The guideline’s February 2024 web update goes a step further and describes the noninferiority result above (AAO Amblyopia PPP, updated February 2024). Our reading of that shift is that it is real but narrow: a legitimate option in specific situations, not a reason to skip the first steps. For what amblyopia therapy looks like session by session, see amblyopia in the eyes of the therapist.

Adherence explains a lot of the disagreement

Look at how much treatment was actually done. In the PEDIG game trials, only 22.2% of the children whose device logs were available — and 13% of the teenagers — completed more than three-quarters of the prescribed play, and median adherence among the children was 46% (Holmes et al., 2016; Manh et al., 2018). A 2026 meta-analysis of 14 randomized trials and 1,110 children found no significant overall difference between video-based dichoptic therapy and patching for acuity or stereoacuity, and reported better results in the children with high adherence to dichoptic therapy, with greater adherence predicting greater improvement. The authors graded all of those findings very low certainty and framed the adherence result as something that may be associated with greater improvement rather than something established (Carvalho Soares Valentim et al., 2026).

Adherence also does not run only one way. In the eye-tracking trial above, children stuck with the device more reliably than with the patch (91% versus 83%) (Wygnanski-Jaffe et al., 2023), so screen-based therapy is not automatically the option children abandon.

Dose realism is an old lesson in this field: six hours of daily patching was no better than two (Repka et al., 2003). The treatment that gets completed usually beats the treatment that looks best on paper.

How we use technology here

This part is our clinical perspective, and we want it clearly labeled as that rather than as a research finding.

• Glasses and standard amblyopia care come first. Refractive correction, then patching or atropine where indicated. Digital tools sit alongside them, not in place of them — see amblyopia and strabismus care.
• Screen-based activities are part of a program, not the program. Lenses, prisms, free-space fusion targets and balance and body-awareness work still make up most of a session in our vision therapy rooms.
• Supervision is the point. In an office session a therapist can see a child tilt their head, close one eye or lean in to compensate, and change the task. A headset at home cannot.
• We track objective measures, not enthusiasm. Acuity, near point of convergence, vergence ranges, accommodative facility and stereoacuity decide whether something stays in a plan.
• Not every child suits a headset. Age, fit, motion sensitivity and seizure history all matter, and some children simply do better with hands-on targets.

Before you buy a headset for home

Ask what condition the device was tested on, what it was compared against, who paid for the study, how many hours a week it expects, and how progress will be measured. If a product claims to cure amblyopia, correct dyslexia or replace a clinical program, treat that as a reason for caution.

Technology has expanded what we can present to two eyes at once. It has not changed the part that decides outcomes: an accurate diagnosis, a plan matched to it, and treatment that is genuinely completed. If your child has been told they have a lazy eye, an eye turn or a focusing problem — or you are an adult still living with one — schedule a comprehensive developmental vision evaluation and we will tell you honestly which tools, digital or not, are likely to help. Schedule an evaluation at our Boynton Beach or Boca Raton office.