The Three Grades of Binocular Vision Explained

Two eyes have a job a single camera never has: send the brain two slightly different pictures and get one world back, with depth in it. Ophthalmology has described that job in three levels for more than a century. The British ophthalmologist Claud Worth sorted binocular single vision into three grades: simultaneous perception, fusion and stereopsis. The classification is still taught because it maps onto what we actually test in the exam room (Bhola, University of Iowa).

Think of the grades as rungs on a ladder rather than three separate skills. Someone can hold the bottom rung and never reach the top, and where they stop shows up on a ball field, behind a windshield and in a book. This article sits inside our complete guide to binocular vision dysfunction.

Key takeaways

• Grade 1, simultaneous perception: the brain registers both eyes’ images at once instead of switching one off.
• Grade 2, fusion: those images blend into one and stay merged while the eyes move - sensory fusion plus a motor effort (Bhola).
• Grade 3, stereopsis: the brain reads the tiny difference between the two images as real depth. It is the most demanding grade, and in our clinical experience the first one to slip when something goes wrong.
• Stereopsis arrives fast - measurable at a mean of 16 weeks (Held et al., 1980) - then sharpens for years (Birch et al., 2008).
• The grades matter most where speed, close distance judgment and sustained near work meet. The evidence is stronger in some of those areas than others, and we say which.

Grade 1: simultaneous perception

The most basic form of binocularity is that both eyes get to speak at once: the visual cortex takes in a stimulus from each eye at the same time, without shutting either down (Bhola).

That sounds automatic, and usually it is. But when the two images disagree - because an eye turns, or one eye’s picture is much blurrier - the brain has a shortcut: suppression. It quietly ignores one eye. Suppression is comfortable, which is exactly why it gets missed. A child who suppresses does not complain of double vision; they simply stop using half of their visual system. Grade 1 is not about depth at all. It is about whether both eyes are still in the conversation.

Grade 2: fusion

Fusion is where two images become one, and it has two halves. Sensory fusion is the brain reading two similar images as a single object. Motor fusion is the muscular work of keeping them lined up while the head turns, the page moves or the eyes shift from the whiteboard to the desk. True grade 2 means fusion with amplitude: not merging the images once, but holding the merge under stress (Bhola).

Amplitude is the part that fails quietly. Fusion holds for the first ten minutes of reading, then slips, and print blurs, doubles or swims. That is the territory of convergence insufficiency and the other eye-teaming problems - and the reason a person can pass a 20/20 acuity check and still be miserable eight inches from a book.

Grade 3: stereopsis

Because the two eyes sit a short distance apart on the face, each retina gets a slightly displaced image, and the brain reads that displacement - binocular disparity - as depth. Depth discrimination is measured in seconds of arc, usually written as arcseconds. An arcsecond is 1/3600th of a degree - 40 arcseconds is about the width of a human hair held at arm’s length. Smaller numbers mean the eyes can tell apart smaller differences in depth, so on a stereo test a lower score is a better score.

Stereopsis is not our only distance cue. Shadow, relative size, motion parallax and overlap all work with one eye. What stereopsis adds is precision, most of it at close range and at speed. Our companion articles on how depth perception develops and stereoscopic vision go further into the mechanics.

How the three grades develop

Depth perception is not present at birth. The American Optometric Association describes the eyes becoming capable of a three-dimensional view at around the fifth month, with depth perception and eye-hand coordination well developed by age 2 (AOA). The American Academy of Ophthalmology adds that a newborn’s eyes often wander or appear to cross for the first two months, but should focus and track together by 3 months (AAO).

Laboratory measurement sharpens the picture. In a study using a preference-looking procedure, stereopsis first became measurable at a mean age of 16 weeks, and infants were resolving disparities of one minute of arc or better by a mean of 21 weeks (Held et al., 1980). That is a fast climb next to the slower maturation of visual acuity.

Refinement then continues for years. In 4,355 children, mean scores on the Randot Preschool stereoacuity test improved from about 100 arcseconds at age 3, to 60 by age 5, to 40 by age 7 (Birch et al., 2008). Those figures belong to that one test. Forty arcseconds is the best score the test offers, and almost every child over 5 with any stereo vision reaches it (Read et al., 2019). Read the numbers as a rough developmental trend, not as a universal scale.

The window in which the system is most shapeable, and most vulnerable, narrows through childhood. Treatment response drops measurably after about age 7: pooled data from four amblyopia trials found children aged 7 to under 13 significantly less responsive than children aged 3 to 7 (Holmes et al., 2011). The classical teaching places the opening of the sensitive period at about 4 months, its peak near age 2, and its close near age 9 (Bhola).

What disrupts them

Anything that stops the two eyes delivering matched, aligned images during that window can cost a grade. The classic teaching sorts the obstacles into three kinds (Bhola):

• Sensory obstacles spoil the picture one eye sends - a cloudy lens or cornea, an uncorrected prescription, a problem in the retina or optic nerve, or long stretches of using one eye alone.
• Motor obstacles stop the eyes pointing at the same place - trouble with the muscles that aim the eyes or the nerves that drive them, which usually shows up as an eye that turns in or out.
• Central obstacles sit further back, in how the brain handles the two images rather than in the eye itself.

What actually caps the grades is not the obstacle alone but how the brain adapts to it. It can suppress, switching one eye’s image off. It can also develop anomalous retinal correspondence. That means the brain quietly re-maps where it treats one eye as pointing, pairing the straight-ahead spot of one eye with an off-center spot of the other. The mismatch then stops being noticeable, which is comfortable but costs depth. Or vision in the weaker eye can fail to develop normally, which is amblyopia.

These are not rare. Pooled across 56 studies, any strabismus affects about 1.93% of people (Hashemi et al., 2019), and amblyopia about 1.44% globally, 2.41% in North America (Fu et al., 2020). Reduced stereoacuity is the most common binocular deficit found in amblyopia under ordinary viewing (Levi et al., 2015). Later in life a sudden misalignment - after a concussion, a nerve palsy or a decompensating phoria - can knock fusion down in an adult who had all three grades for decades. See double vision in children and our amblyopia and strabismus page.

Why the grades matter

On the field

In 252 professional baseball players, scores on a sensorimotor test battery predicted on-base percentage, walk rate and strikeout rate, after adjusting for age, position and league (Burris et al., 2018). On the other side of the same coin, the stereo deficit seen in amblyopia is described as affecting motor coordination, sport and mobility (Levi et al., 2015). The honest caveat: those studies measure ability, not the effect of training. A review of digital sports vision training describes the field as resting on one premise: that demanding visual, perceptual, cognitive and oculomotor exercises can improve how well an athlete processes and responds to what is seen. The same review can only appraise the supporting research where such evidence exists (Appelbaum & Erickson, 2018).

Behind the wheel

Here the literature is genuinely mixed, and we will not pretend otherwise. A review of driving and binocularity found raised crash rates among commercial drivers with reduced stereopsis, but no such association in non-commercial drivers. The same review cites experimental data in which reduced-stereopsis drivers braked earlier and crashed less (Dakroub et al., 2022). What shows up consistently is sudden change: when binocular vision was artificially degraded in a simulator, driving became unsafe in the most complex traffic environments (Molina et al., 2021). Reading those findings together, our clinical view at Vision & Learning Center is that long-standing reduced stereopsis is usually well compensated, while a sudden change in fusion is the one that needs prompt attention.

At the desk

Pegboard and bead-threading performance tracked stereoacuity in 143 people aged 10 to 30, with the normal-stereoacuity group best at every task (O’Connor et al., 2010). Among 4- and 5-year-olds, the widest early-literacy gap fell to one subgroup: children with uncorrected moderate farsightedness whose near stereoacuity measured 240 arcseconds or worse. They scored 8.6 points below children who needed no correction at all (Kulp et al., 2016). That study compared uncorrected farsighted children against children who needed no glasses. So it shows an association with uncorrected farsightedness in a group whose near vision was also under strain. It is not evidence that treating stereoacuity by itself raises literacy scores.

Reading deserves a careful sentence. In 5,822 children aged 7 to 9, four out of five with severe reading impairment had entirely normal ophthalmic function. The study found no association with strabismus, refractive error, amblyopia, convergence or accommodation, though stereoacuity worse than 60 arcseconds was somewhat more common among the poor readers (Creavin et al., 2015). And when vision therapy for convergence insufficiency was tested against placebo therapy for reading comprehension, it produced 3.7 points versus 3.8 (CITT-ART, 2019). Our clinical view at Vision & Learning Center is that treating eye teaming makes near work more comfortable and sustainable. It is not a reading intervention; reading instruction belongs to teachers and reading specialists.

How the grades are measured

In practice we assess the grades in layers. Cover testing and alignment measures come first, then Worth 4-dot and other suppression checks for grade 1, then vergence ranges and near point of convergence for grade 2 fusion amplitude, and a stereo test for grade 3. A stereo test alone is a weak screen: the Randot Preschool test is highly specific (over 95%) but its sensitivity for detecting visual problems is poor, under 50% (Read et al., 2019). Acuity charts have the same limitation. The AOA’s pediatric guideline reports that distance acuity testing alone missed 75.5% of children later found to have binocular and oculomotor problems on a full examination. That guideline recommends comprehensive exams at 6-12 months, at least once between ages 3 and 5, and annually from first grade (AOA, 2017).

What can be improved

It depends on which grade is limited and why. The strongest evidence covers one specific eye-teaming problem in children whose eyes are otherwise straight: symptomatic convergence insufficiency. In a 221-child randomized trial, 73% of children aged 9 to 17 were successful or improved after 12 weeks of office-based vergence and accommodative therapy, against 35% on office placebo (CITT, 2008). Most of those children - 164 of the 221 - also had a focusing problem. Their amplitude of accommodation, meaning the range over which they could pull near print into focus, gained 9.9 diopters with office-based therapy against 2.2 with placebo (Scheiman et al., 2011). A focusing problem on its own is a different and less settled question. A review of accommodative insufficiency found no high-quality evidence for a standard treatment protocol. Vision therapy and low plus lenses are each supported by separate studies, and the two have never been compared head to head (Hussaindeen & Murali, 2020).

Amblyopia and constant strabismus are harder, and timing matters. Pooled data from four trials found children aged 7 to under 13 significantly less responsive to amblyopia treatment than children aged 3 to 7, although some older children still improved markedly (Holmes et al., 2011).

Fine stereopsis is harder to restore after strabismic amblyopia (an eye turn) than after anisometropic amblyopia (a large difference in prescription between the eyes), and training the weaker eye on its own rarely helps those patients. The same review is more hopeful about newer methods. Strabismic amblyopes did better with dichoptic training, where each eye is shown a different piece of one shared task, than with training one eye alone - and better still with training aimed directly at depth (Levi et al., 2015). We would rather say that plainly than promise a grade we may not be able to give back.

If reading is uncomfortable, if depth judgment seems off, if an eye drifts, or if something changed suddenly, the three grades are exactly what a comprehensive developmental vision evaluation is built to measure. Schedule an evaluation at our Boynton Beach or Boca Raton office and we will tell you which rungs are solid and which need work.