Does Vision Therapy Work? An Honest Evidence Summary

Vision therapy may be the most argued-about treatment in eye care. Search for it and you will find optometry practices describing transformed readers, and ophthalmology organizations calling it unproven. Both sides are quoting real sources. Neither, on its own, is telling you the whole story.

This page is our attempt at the whole story. We are a vision therapy practice, so we have an obvious stake in the answer, and we are going to tell you where the research does not support us as clearly as where it does.

Key takeaways

• The strongest evidence in the field supports office-based vergence and accommodative therapy for symptomatic convergence insufficiency in children, tested against placebo in randomized trials.
• The same research group’s later, larger trial found no reading benefit and no symptom advantage over placebo, which is an important limit on what anyone should claim.
• Evidence for focusing disorders and for adults is real but thinner: smaller studies, fewer controls. Post-concussion convergence insufficiency now has a 104-participant randomized trial behind it, though it compared starting therapy immediately with starting it six weeks later, not with a sham.
• Vision therapy is not a treatment for dyslexia, ADHD or amblyopia. Claims that it is go past the data.
• The most useful thing you can do is ask a provider what they are treating, how they will measure it, and when they would stop.

Three different voices, and why they disagree

Almost every argument about vision therapy collapses into confusion because three separate voices get mixed together.

The research literature answers narrow questions: did this specific treatment, for this specific diagnosis, beat a placebo in a randomized trial? That is a much smaller question than “does vision therapy work.”

Professional bodies issue positions, and they genuinely disagree. The American Academy of Pediatrics, American Academy of Ophthalmology, AAPOS and AACO hold jointly that vision problems can interfere with learning but do not cause dyslexia, and that evidence does not support eye exercises or behavioral vision therapy as treatments for learning disabilities (AAP/AAO/AAPOS/AACO, 2009); that statement was reaffirmed in 2014 with the language that such treatments “are not endorsed and should not be recommended” for educational performance (AAO et al., 2014). The American Optometric Association and American Academy of Optometry hold that vision problems can and often do interfere with learning, and that people at risk should be evaluated as part of multidisciplinary care (AAOpt/AOA, 1997).

A practice’s clinical perspective is the third voice, and it is the one you are reading here. It is worth something, but it is not evidence, and we will label it as ours when we get to it.

Where the evidence is strongest: convergence insufficiency in children

Convergence insufficiency is difficulty holding the eyes turned inward comfortably at reading distance. It is common: about 13% of fifth- and sixth-graders in a US school-based study met a high-suspect or definite definition (Rouse et al., 1999).

The Convergence Insufficiency Treatment Trial randomized 221 children aged 9 to 17 to office-based vergence/accommodative therapy with home reinforcement, home computer therapy, home pencil push-ups, or an office-based placebo. After 12 weeks, 73% of the office-based therapy group were successful or improved, versus 35% of the placebo group, 43% for pencil push-ups and 33% for home computer therapy (CITT Study Group, 2008). Among children who became asymptomatic, 84.4% of the office-therapy group were still asymptomatic a year later (CITT Study Group, 2009).

A Cochrane network meta-analysis pooling 12 trials and 1289 participants reached the same conclusion: office-based therapy with home reinforcement was more effective than placebo in children, with a risk ratio of 3.04 (Scheiman et al., 2020).

Two honest footnotes. First, 35% of the placebo group also improved, so the real gap over placebo is roughly 38 percentage points, not 73. Second, pencil push-ups alone, which are still handed out as a home fix, did not beat placebo.

What the follow-up trial found

CITT-ART is the most important study for keeping expectations straight. In the 311 children it randomized, clinical signs improved more with therapy than placebo: near point of convergence improved 10.4 cm versus 6.2 cm, and positive fusional vergence 23.2 versus 8.8 prism diopters. But symptom scores improved by similar amounts in both groups (CITT-ART, 2019). And in the companion reading analysis of 310 of those children, therapy produced a 3.7-point gain in reading comprehension versus 3.8 points for placebo (CITT-ART, 2019).

We link to that trial deliberately, and we walk through it in more depth in what the CITT and CITT-ART trials actually found. Any practice that cites the 2008 result and never mentions 2019 is showing you half the file.

Where the evidence is moderate or limited

Focusing (accommodative) problems. Within CITT, 164 of 221 children (74%) also had accommodative dysfunction, and office-based therapy raised accommodative amplitude by 9.9 D versus 2.2 D with placebo (Scheiman et al., 2011). For isolated accommodative insufficiency, though, a review found reported prevalence ranging from under 1% to 61.6% depending on criteria, and no high-quality evidence for a standard treatment; plus lenses and vision therapy have each shown benefit separately but have never been compared head to head (Hussaindeen and Murali, 2020).

After a concussion. Vision problems are genuinely common in children referred with concussion: 69% of 100 adolescents in one concussion program had a vision diagnosis (Master et al., 2016), and in a separate sample of 113 adolescents assessed 4 to 12 weeks after injury, 70% had an oculomotor diagnosis, with routine physician screening missing many (Scheiman et al., 2021). The AAP and AAO clinical report agrees these problems matter, while noting most children recover on their own by four weeks and recommending a vision-specific examination for those who do not (Master et al., 2022).

Treatment evidence is thinner than diagnosis evidence, but it is no longer absent. Most children get better without therapy: in one series, 46% of children with abnormal convergence after concussion recovered with standard care and only 13% went on to office-based vision therapy (Storey et al., 2017). The high success rates quoted in practice marketing come from uncontrolled private-practice records (Gallaway et al., 2017). A small randomized trial of 50 people found home Brock string work improved the convergence measurement more than usual care, though it did not test whether recovery was shortened (Trbovich et al., 2025).

The concussion-specific randomized trial of office-based therapy is CONCUSS. It randomized 104 people aged 11 to 25 who still had symptoms and convergence insufficiency 4 to 24 weeks after injury, either to immediate office-based vergence and accommodative therapy or to the same therapy delayed by six weeks. At six weeks, 46 of 52 (88%) in the immediate group were successful or improved on the primary outcome, a composite of near point of convergence and positive fusional vergence, versus 4 of 52 (8%) in the delayed group, and symptoms improved in 79% versus 13% (Alvarez et al., 2026). A secondary analysis of the same trial found focusing amplitude improved by roughly 4.7 to 5.1 diopters with immediate therapy versus about 1.0 with waiting (Scheiman et al., 2026). We treat that as real evidence, and we keep three limits in view: the comparison group waited rather than receiving a sham therapy, so expectation effects are not ruled out; the endpoint was eye-teaming and focusing measurements, not concussion recovery or return to play; and once both groups had completed all 16 sessions, their outcomes were equivalent, which is why the authors describe immediate therapy as shortening the symptomatic period rather than changing the destination. No trial has yet used overall concussion recovery or return to play as its primary endpoint, and none has used a sham comparison in concussed patients. We go through this literature in more detail in vision problems after a concussion.

Adults. In a trial of 46 adults aged 19 to 30, only the office-based therapy group significantly improved convergence measurements, but 42% reached symptom-elimination criteria versus 31% of placebo, and more than half remained symptomatic (Scheiman et al., 2005). Cochrane concluded the adult evidence is less clear than the pediatric evidence (Scheiman et al., 2020).

Where the evidence is negative, or simply missing

Reading and dyslexia. CITT-ART is negative for reading. A population cohort of 5822 children found no association between severe reading impairment and strabismus, refractive error, amblyopia, convergence or accommodation, and four of every five poor readers had entirely normal ophthalmic function (Creavin et al., 2015). A smaller clinical study did find more visual deficits in children with dyslexia: 79% had a deficit in at least one domain of visual function versus 33% of typical readers. Broken out, that was accommodation 55% versus 9% and ocular motor tracking 62% versus 15%, while the difference in vergence deficits (34% versus 15%) was not statistically significant. The authors themselves wrote that the cause and clinical relevance of these deficits are uncertain (Raghuram et al., 2018). And where vision therapy has been given to children with learning disorders, the binocular vision measures improved, but reading was never measured as an outcome in that study, so it tells us nothing either way about reading (Hussaindeen et al., 2018). The pediatric technical report is right that the treatment for dyslexia is evidence-based reading instruction (Handler and Fierson, 2011).

Attention and ADHD. Convergence insufficiency and ADHD co-occur about three times more often than expected (Granet et al., 2005), and in a national survey of 75,171 children, ADHD was reported in 15.6% of those with vision problems versus 8.3% without (DeCarlo et al., 2016). These are associations. There is no trial showing vision therapy treats ADHD, and we do not claim one.

Amblyopia. The AAO Preferred Practice Pattern is blunt: antisuppression, vergence, accommodation, eye-hand coordination activities and perceptual learning are often prescribed as adjuncts, but “there is no convincing evidence for treatment success with these activities for amblyopia” (AAO PPP Panel, 2023). The AAO’s separate review of binocular treatment found no level I evidence that it can substitute for patching or glasses (Pineles et al., 2020); large trials found a binocular iPad game inferior to patching (Holmes et al., 2016) and no better than a placebo game (Gao et al., 2018). A newer dichoptic digital therapeutic did beat glasses alone, 1.8 lines versus 0.8 (Xiao et al., 2022), but it was not compared with patching. Glasses, patching and atropine remain the treatment.

Sports vision. The rationale is plausible and the evidence is uneven (Appelbaum and Erickson, 2018). The most-quoted study raised one college team’s batting average from 0.251 to 0.285 with no randomization and no control for coaching or roster changes (Clark et al., 2012). Interesting, not proof.

Myopia, blue light and light therapy. Time outdoors protects against myopia onset but does not slow progression in eyes that are already myopic (Xiong et al., 2017). Blue-light-filtering lenses may not reduce eye strain at all (Singh et al., 2023). And syntonic phototherapy, which some optometric practices offer, was reviewed across eight studies with GRADE certainty rated very low for every outcome and no consistent evidence of effect (Cervera-Sanchez et al., 2023).

What the critics say, and where they are right

The most thorough independent critique came from a UK academic optometrist who examined ten categories of behavioral vision therapy and concluded that the large majority are not evidence-based, singling out convergence insufficiency, yoked prisms in neurological patients, and rehabilitation after brain injury as the areas with real support (Barrett, 2009).

That is a fair description of the field, and it lines up with everything above. The critics are right that a lot of what gets marketed under the vision therapy umbrella has never been tested against a placebo; that success rates quoted from uncontrolled case series are not comparable to trial results; and that claims about reading, grades and attention have been tested and did not hold up. Where we part company is the further step some critics take, that because the learning claims failed, the underlying binocular and accommodative problems are not worth diagnosing or treating. The CI trials say otherwise.

How to judge any provider’s claims

• Ask what is being treated. A named diagnosis with measurable signs, not “visual processing” in general.
• Ask how it will be measured. Near point of convergence, fusional ranges, accommodative amplitude and facility, and a validated symptom survey are numbers you can see move.
• Ask about the stopping rule. A program that has no point at which it would say “this is not working” is not a clinical plan.
• Watch the outcome being promised. Comfort at near, fewer headaches and less avoidance are defensible. Reading level, grades and attention are not.
• Be careful with percentages. A “90% success rate” from a practice’s own completed cases is not the same kind of number as a trial result.

Our clinical perspective

Here is where we tell you plainly what is ours rather than the literature’s. We think the honest position for a vision therapy practice is narrow and specific: we look for binocular and accommodative problems that make near work uncomfortable, we treat the ones the evidence supports treating with office-based vision therapy, and we say so when a child’s difficulty is a reading, language or attention problem that needs a different professional. Vision therapy is one tool in a multidisciplinary picture, not the explanation for everything.

That also means we tell families when they are not a good fit, which we cover in who is a good fit for vision therapy and in our plain-English guide to what vision therapy is. If you are weighing the commitment, the cost and time factors matter as much as the evidence does.

Where this leaves you

If your child or you have symptoms at near, double vision, headaches after reading, losing place, avoiding homework, then the right next step is not to decide about vision therapy. It is to find out whether a measurable binocular or focusing problem is there at all. That is what a comprehensive developmental vision evaluation is for, and the answer may well be that nothing in the visual system explains the difficulty.

We would rather tell you that than sell you a program. If you would like an evaluation at our Boynton Beach or Boca Raton office, Schedule an evaluation.