How Big Are the Letters on an Eye Chart?

The largest letter on a standard eye chart stands roughly 88 millimeters tall, about three and a half inches, when viewed from the standard testing distance of 20 feet (6 meters). The smallest line you need to read for “normal” 20/20 vision is about 8.7 millimeters tall at that same distance. Every line in between follows a precise geometric scaling, and the system behind those sizes dates back to a 19th-century Dutch ophthalmologist whose basic logic still governs how your vision gets measured today.

The 5-Arcminute Rule

Hermann Snellen introduced the first standardized eye chart in 1862, and his key insight was deceptively simple: design each letter so that the overall character covers a visual angle of 5 arcminutes when seen from the correct distance, with each stroke or gap within the letter spanning 1 arcminute. An arcminute is one-sixtieth of a degree. At 20 feet, 1 arcminute translates to about 1.75 millimeters on the chart, so a 5-arcminute letter works out to roughly 8.7 millimeters tall. That is the 20/20 line.¹

The other lines scale up from there. The 20/40 line, which represents half the resolving ability of 20/20, has letters twice as tall: about 17.5 millimeters. The 20/100 line is five times the height of the 20/20 letters, around 43.5 millimeters. And the big “E” at the top of most Snellen charts sits on the 20/200 line at roughly 88 millimeters, ten times the 20/20 size. Every step between those lines represents a proportional change in letter height, not a fixed one. The chart is a ladder of ratios, not of evenly spaced rungs.

The “20/200” fraction itself means something concrete: a person who can only read that top letter at 20 feet is seeing what someone with normal vision could read from 200 feet away. Conversely, if you can read the 20/15 line, your eyes resolve detail that a normal-sighted person would need to be 15 feet closer to match. The fraction is a comparison of distances, not a percentage of perfect vision, and treating it like a simple fraction can mislead. Early commentary on the Snellen system noted that the notation was never meant to express a proportion of “full vision” and that using it that way gives a false impression.²

Why the Spacing Between Lines Matters as Much as the Letters

If you have ever squinted at an eye chart, you may have noticed that reading isolated letters is easier than reading a line of them. That is partly because of a phenomenon called visual crowding: when letters sit close together, your brain has a harder time distinguishing one from its neighbors, especially in your peripheral vision. Research on reading and letter spacing has confirmed that tightly spaced characters constrain how efficiently you process what is next to whatever you are fixating on.³

On a well-designed chart, the gaps between letters and between rows are standardized to keep crowding consistent across every line. The ETDRS chart, which has become the gold standard in research settings, uses uniform spacing proportional to letter size: each letter is separated from its neighbor by one letter width, and rows are spaced by one letter height. Snellen charts, on the other hand, often crowd more letters onto the smaller lines and fewer onto the larger lines. That inconsistency is one reason two people can get different scores depending on which chart they read.

Snellen Charts Versus ETDRS Charts

Most people picture the classic Snellen chart when they think of an eye exam: a big E at the top, progressively smaller lines below, and an uneven number of letters per row. It works reasonably well for everyday screening, but it has some quirks. The number of letters per line varies, the size steps between lines are not uniform, and the letter selection is not controlled, so some lines are harder than others purely because of which letters appear on them.

The ETDRS chart, developed in the 1980s for a large diabetic retinopathy study, addressed those issues. It uses five letters on every line, a consistent logarithmic size progression (each line is about 0.1 log units smaller than the one above it), and a restricted set of letters chosen for roughly equal difficulty. A direct comparison of the two chart types found that patients scored an average of about six and a half letters better on the ETDRS chart, and the gap widened for people with poorer vision. In a subgroup with acuity worse than 20/200, the difference reached about ten letters.⁴

That is not a trivial difference if your treatment plan or eligibility for a clinical trial hinges on a specific acuity score. A computer simulation comparing the two charts found that the ETDRS design was also more reproducible: test-retest variability was roughly half what it was on the Snellen chart.⁵ In practical terms, if you are retested on a Snellen chart and your score shifts by a line, it is hard to know whether your vision actually changed or the chart just behaved differently. With an ETDRS chart, a shift of that size is more likely to reflect a real change.

Despite all that, most private-practice eye exams still use Snellen-style charts. The ETDRS chart takes longer to administer, and for a routine visit where the question is just “does this person need new glasses,” the Snellen chart gets the job done. Research settings and clinical trials overwhelmingly favor ETDRS because precision matters more there.

What About Charts That Do Not Use Letters?

Not everyone being tested can read the Latin alphabet. Young children, people who are illiterate, and patients who use non-Latin writing systems all need an alternative. Several exist, and they follow the same underlying sizing logic: the symbol must subtend 5 arcminutes at the target acuity level.

The Tumbling E chart uses a capital E rotated in four directions. The patient points or says which way the “legs” face. The Landolt C chart uses a ring with a gap in one of several positions, and the patient identifies where the gap is. A study comparing these to standard letter charts found that for uncorrected vision, both tracked closely with letter-based acuity, with correlation values above 0.93. When vision was corrected with lenses, the agreement was weaker, particularly for the Landolt C, which tended to produce slightly lower acuity scores.⁶

For children, Lea Symbols charts use four simple shapes (a circle, a square, a house, and an apple) that even three-year-olds can name or match. In a hospital setting, Lea Symbols could be tested on both eyes in about 77% of young children, compared to only 48% who could complete a Landolt C test.⁷ The trade-off is that Lea Symbols tend to give slightly more generous acuity scores. One study found that children scored about half a line better with Lea Symbols than with standard ETDRS letters.⁸ A comparison in school-age children and adults showed the gap could be as large as one full logMAR line, and the discrepancy grew in people with worse vision.⁹ The shapes are simply easier to distinguish than letters because they have fewer features that can be confused with each other.

How Non-Latin Scripts Change the Size Requirements

The 5-arcminute standard was built around the Latin alphabet, where most letters can be constructed inside a 5-by-5 grid. But writing systems with more complex characters need more visual detail to be recognized. Traditional Chinese characters are a striking example. A study comparing equivalent legibility between traditional Chinese characters and ETDRS optotypes found that readers with normal vision needed the Chinese characters to be about 1.7 to 2.0 times larger than the corresponding ETDRS letter size before they could reliably read them.¹⁰

That enlargement factor has real implications for near-vision testing and reading prescriptions. If an eye exam uses Latin-letter charts to determine reading acuity, a patient whose daily reading is in Chinese, Arabic, Hindi, or another visually complex script may find that their prescribed correction falls short of comfortable reading. The mismatch is geometric, not cultural. More strokes per character means more fine detail the eye must resolve, which means the critical size for reading in that script is inherently larger.

Digital Screens and the Pixel Problem

Printed eye charts are being replaced by digital screens in many clinics. A monitor or tablet can display any letter at any size, randomize the presentation to prevent memorization, and adjust contrast or lighting on the fly. But the move to digital introduces a constraint that printed charts never had: pixels.

A printed letter can have perfectly smooth edges. A letter on a screen is made of discrete squares, and when the letter gets small enough, those squares become visible as jagged stairstepping. At the 20/20 level, each stroke in the letter spans just a few pixels on a typical monitor. Research into how pixelation affects acuity measurement has produced a practical guideline: for clean, filtered (anti-aliased) letters, the largest pixel in the display should be no more than 0.6 times the width of the smallest stroke being tested, and for unfiltered letters, no more than 0.35 times that stroke width.¹¹ In practice, this means testing 20/20 vision on a standard office monitor requires at least a high-definition display, and testing sharper-than-normal acuity on a tablet is borderline.

Modern digital acuity systems address this by specifying minimum screen resolution and calibrating both the pixel size and the luminance of the display. One validated digital chart used a screen resolution of 1280 by 1024 pixels, with optotype luminance of 2.6 candelas per square meter on a 135 candela background, and a maximum contrast ratio of 500 to 1.¹² When commercially available tablets have been tested for clinical use, researchers found that each device had a different luminance range and gamma curve, and individual calibration was essential before using any of them for precise contrast or acuity testing.¹³

This is worth knowing if you have ever taken a “vision test” on your phone or laptop at home. Without calibration for your specific screen’s size, resolution, brightness, and distance, the result is little more than a guess. An online test might tell you that you read a letter equivalent to 20/30, but if the screen rendered that letter a few pixels too large or the viewing distance was off by even a foot, the measurement is meaningless.

Near-Vision Charts and How They Differ

Eye charts for near vision, the kind you hold at arm’s length or at reading distance, follow the same angular-size logic but at much shorter distances, typically 40 centimeters (about 16 inches). Because the testing distance is so much shorter, the physical letters are dramatically smaller. A 20/20-equivalent letter at 40 centimeters is only about 1.16 millimeters tall. You can barely see the stroke detail with the naked eye at that scale, which is the whole point: it tests whether you can.

Several competing notation systems exist for near charts. The Jaeger system (J1, J2, J3, and so on) is the oldest and the least standardized, with letter sizes that vary between manufacturers. The N-notation system (N5, N6, N8) uses metric point sizes and is more consistent. The Rosenbaum pocket chart uses Snellen equivalents. And ETDRS near charts scale the same logarithmic design down to the shorter distance. A comparative validation study of these handheld charts found that converting between systems requires careful measurement of the actual character heights on each chart, because nominally equivalent sizes can differ in practice depending on the typeface and manufacturer.¹⁴

If you have been told you have good distance vision but struggle to read fine print, a near-vision test using one of these charts is what determines whether you need reading glasses or a multifocal prescription. The sizes involved are small enough that even the printing quality of the chart matters.

Testing Low Vision at the Other Extreme

Standard charts bottom out at 20/200, but plenty of people have vision worse than that. Testing those patients requires either larger optotypes or shorter viewing distances. Low-vision charts like the Feinbloom chart use a series of numbers scaled up to very large sizes, and they allow testing at distances as close as a few feet. The LEA Numbers Low Vision Book takes a similar approach with simpler symbols. Both are evaluated against the ETDRS chart as a benchmark, with the total number of correctly identified optotypes recorded and converted to a common logarithmic scale so that results from different charts can be compared.¹⁵

For someone with 20/400 vision, the letter they can just barely read at 20 feet would need to be about 175 millimeters tall, roughly seven inches. At 20/800, you are looking at a letter the size of a paperback novel’s cover. At some point the chart runs out of wall space, and that is when the examiner walks you closer to the chart instead of making the letters bigger. Moving from 20 feet to 10 feet effectively doubles the visual angle of every letter, turning the 20/200 line into a 10/200 test (equivalent to 20/400).

What Limits How Small a Letter You Can See

The 5-arcminute standard for 20/20 vision was not chosen arbitrarily. It is close to the theoretical resolution limit imposed by the anatomy of the retina. The cone cells packed into the center of your fovea, the tiny region of sharpest vision, are each about 3 micrometers across. Two adjacent points of light have to land on separate cones with at least one unstimulated cone between them before you can perceive them as distinct. A theoretical analysis assuming a 3-micrometer cone diameter and a 4-millimeter pupil estimated the finest resolution the human eye can achieve at about 0.05 diopters of spherical change, corresponding to a visual angle right around 1 arcminute for the gap or stroke within a letter.¹⁶

Some people test better than 20/20. Young adults with excellent optics and dense foveal cone packing routinely achieve 20/15 or even 20/10 in clinical settings. At 20/10, the letter is only about 4.4 millimeters tall at 20 feet, and the strokes within it span half an arcminute. That approaches the hard ceiling set by photoreceptor spacing and the diffraction of light through the pupil. Beyond that, no amount of squinting or motivation helps; the physics of the eye will not allow it.

How Testing Distance Changes Everything

All the letter sizes discussed above assume the standard 20-foot (6-meter) testing distance. But many clinics do not have rooms that long. A common workaround is to use a mirror to double the effective path length: the chart sits behind the patient, who reads its reflection in a mirror on the far wall. Another approach is to use a chart calibrated for a shorter distance, often 4 meters or 3 meters, with the letter sizes scaled down proportionally. A 20/20 letter at 4 meters is only about 5.8 millimeters tall instead of 8.7 millimeters, because the viewing distance is two-thirds of the standard.

The ETDRS protocol actually accounts for this by including a procedure to move the patient from 4 meters to 2 meters if they cannot read enough letters at the standard distance. This effectively doubles the angular size of every letter on the chart without changing the chart itself. It is a simple and elegant solution, but it introduces a potential discrepancy with Snellen-based results, which typically do not have a built-in distance-halving step. That methodological difference is one more reason why acuity scores from different tests do not always line up neatly.