logMAR vs Snellen: two ways to score visual acuity

If you have ever compared a vision result from an eye clinic against one from a research study — or tried to make sense of a number like "0.1 logMAR" next to the familiar "20/20" — you have run into the two competing ways the eye-care world scores visual acuity. One is the fraction most people know: 20/20, 20/40, 20/200. The other is a decimal that looks like nothing in particular until someone explains it: logMAR. They measure the same underlying thing — the smallest high-contrast letters you can reliably read — but they score it on different scales, and the difference matters more than it looks.

This post explains what each scale is, why vision researchers largely abandoned the Snellen fraction for clinical trials, and why none of this changes the deeper limitation both scales share: a single high-contrast acuity number, in either format, is a narrow snapshot of vision.

TL;DR. Snellen acuity (20/20, 20/40) is a ratio comparing your performance to a "normal" reference distance. logMAR is the base-10 logarithm of the minimum angle of resolution — 0.00 logMAR equals 20/20, and every 0.1 logMAR step is one line on a well-designed chart. Researchers prefer logMAR because its charts have equal letters per line, equal spacing, and a linear scale you can average and do statistics on. Snellen charts don't. Both scales still only measure high-contrast letter reading, which is why contrast sensitivity captures things acuity misses.

What the Snellen fraction actually means

The Snellen chart, designed by the Dutch ophthalmologist Herman Snellen in 1862, is the wall of shrinking letters everyone pictures. Your score is written as a fraction: the top number is your testing distance (20 feet in the US, or 6 metres elsewhere — hence 6/6), and the bottom number is the distance at which a person with "normal" vision could read the same line.

So 20/40 means: the smallest line you could read at 20 feet is one that a reference observer could have read from 40 feet. Bigger bottom number, worse acuity. 20/200 — the US threshold for legal blindness in the better eye with best correction — means the smallest line you managed at 20 feet is one a reference observer reads from 200 feet.

The fraction is intuitive as a description. Its problems are structural, and they only become obvious when you try to measure change precisely or compare across charts.

Why researchers moved to logMAR

logMAR stands for the logarithm of the Minimum Angle of Resolution. The minimum angle of resolution (MAR) is the size, in minutes of arc, of the smallest detail you can resolve. A person reading the 20/20 line resolves detail subtending about 1 minute of arc, so their MAR is 1.0 and the base-10 logarithm of 1.0 is 0.00. That is why 0.00 logMAR = 20/20.

The scale runs the intuitive-feeling way backwards: lower is better, and negative numbers are better than 20/20. Reading the 20/16 line is about −0.1 logMAR; 20/40 is +0.3; 20/200 is +1.0. Each 0.1 logMAR corresponds to one line on a properly built chart, and each line changes letter size by a constant ratio (about 1.26×, the fifth root of 10).

The scale itself is only half the story. The bigger reform was the chart. In 1976, optometrists Ian Bailey and Jan Lovie-Kitchin published a set of design principles that fixed the structural flaws of the Snellen layout (Bailey & Lovie, 1976). Their chart, and the ETDRS chart derived from it for the Early Treatment Diabetic Retinopathy Study (Ferris, Kassoff, Bresnick & Bailey, 1982), became the research standard. The design rules were deceptively simple:

• Equal letters per line — five on every row, so the task is the same difficulty-per-letter top to bottom. A standard Snellen chart has one big letter on top and many small ones at the bottom, so "reading a line" means something different at each size.
• Equal, proportional spacing between letters and rows, so crowding (the interference between adjacent letters) is constant. On a Snellen chart, the big top letters have huge gaps and the small bottom ones are cramped, which changes the test in mid-chart.
• A logarithmic size progression — each row a fixed ratio smaller — so the steps are perceptually even and the scale is linear in logMAR.
• Letters balanced for legibility across rows, so no line is unusually easy or hard because of which letters it happens to contain.

The payoff is measurement quality. Because every line contributes equally and the scale is linear, you can score letter-by-letter (each of the five letters on a line is worth 0.02 logMAR), average scores across eyes or visits, compute standard deviations, and run parametric statistics. You cannot cleanly average Snellen fractions — what is the mean of 20/20 and 20/200? The arithmetic is meaningless because the underlying scale is not linear. For a clinical trial trying to detect a two-line change, this is the difference between a usable outcome measure and a noisy one.

They don't always agree in the real world

You might assume the two scales are interchangeable — just convert the fraction to a decimal. In controlled conditions they line up. In routine clinics they drift apart, because the charts and procedures differ.

Kaiser, in a prospective study in a working retinal practice, compared Snellen acuity measured the usual clinical way against standardised electronic ETDRS testing and found the two methods did not produce equivalent scores; the discrepancy was large enough to matter when a treatment decision hinges on a line or two of change (Kaiser, 2009). The reasons are mundane but real: Snellen testing lets patients stop at the first line they miss a letter on, uses inconsistent letter counts per line, and varies with how forgiving the examiner is about near-misses. ETDRS forces every letter, scores each one, and uses a fixed termination rule. Same eye, same day — different number.

The practical upshot: a Snellen acuity from your optometrist and a logMAR acuity from a research protocol are not directly comparable to the second decimal place. A rough conversion is fine for orientation; treating them as identical measurements is not.

Neither scale measures the vision you use most

Here is the point both scales quietly share, and the reason this whole comparison matters for us. logMAR fixed the measurement problems of the Snellen chart. It did nothing to widen what is measured. Both are high-contrast, black-on-white letter tests. They ask one question extremely well: what is the smallest sharp-edged, maximum-contrast symbol you can identify?

That is not the visual world. Faces are low contrast. Roads at dusk are low contrast. Newsprint, fog, a grey cat on a grey carpet — low contrast. A person can score a crisp 20/20 (0.00 logMAR) and still struggle with all of those, because the ability to see reduced-contrast patterns is a partly separate function of the visual system: contrast sensitivity. We wrote a whole post on why a single high-contrast acuity number is a misleading summary of vision, and a primer on what contrast sensitivity actually measures that fills in the gap acuity leaves.

Contrast sensitivity has its own measurement history that parallels the logMAR reform. The Pelli-Robson chart (Pelli, Robson & Wilkins, 1988) did for contrast what Bailey and Lovie did for acuity: a principled, well-spaced letter chart with a logarithmic step size, this time varying contrast rather than size. More recent adaptive methods like the quick CSF estimate a whole sensitivity curve efficiently (Lesmes, Lu, Baek & Albright, 2010). The lesson across both domains is the same: the scale you choose determines how much you can trust small differences. And the way clinics decide which chart and method to use is worth understanding, which is why we compared Pelli-Robson, FACT and qCSF directly.

What this does and does not mean

Note. A visual acuity score — in Snellen or logMAR — is a measure of high-contrast letter reading. It is not a diagnosis, and neither format tells you why an acuity is reduced. A contrast sensitivity test is a screening signal of a different visual function; it also does not diagnose anything and does not replace an eye exam.

Converting between the scales is fine for rough orientation (0.00 logMAR ≈ 20/20, +0.3 ≈ 20/40, +1.0 ≈ 20/200). It is not reliable to the precision a clinical decision needs — for that, the same chart and procedure has to be used both times.

A normal acuity on either scale does not rule out a vision problem. Early cataract, glaucoma, and macular disease can leave high-contrast acuity intact while degrading contrast, glare tolerance, or the visual field.

Frequently asked questions

Is logMAR better than Snellen?

For measurement precision, research, and tracking change over time — yes, meaningfully. logMAR charts have equal letters per line, constant spacing, a linear scale you can average and analyse statistically, and letter-by-letter scoring. Snellen is faster and more familiar for a quick clinical screen, which is why it survives in practice. "Better" depends on whether you are screening or measuring.

What is 20/20 in logMAR?

20/20 is 0.00 logMAR. 20/40 is +0.3 logMAR (three lines worse). 20/16 is about −0.1 logMAR (one line better than 20/20). Every 0.1 logMAR is one line on a standard chart; every 0.02 logMAR is one letter.

Why is lower logMAR better?

Because logMAR is the logarithm of the minimum angle of resolution — the size of the smallest detail you can resolve. Smaller detail resolved means sharper vision, and a smaller angle means a smaller (eventually negative) logarithm. So the scale runs opposite to intuition: 0.00 is 20/20, and better-than-normal vision goes negative.

Can I convert my Snellen score to logMAR exactly?

Only approximately. The formula logMAR = log₁₀(denominator ÷ numerator) gives the nominal value (e.g. 20/40 → log₁₀(2) ≈ 0.30). But because Snellen and logMAR charts differ in letters-per-line, spacing, and scoring, two real measurements of the same eye can disagree by a line or more (Kaiser, 2009). Use the conversion for orientation, not for tracking small changes.

Does a good acuity score mean my vision is fine?

Not necessarily. Acuity in either scale only measures high-contrast letter reading. It can be normal while contrast sensitivity, glare recovery, or peripheral vision are reduced. This is exactly why a contrast sensitivity screen measures something acuity leaves out.

Take the test

Acuity charts tell you how small you can see. A contrast sensitivity test tells you how faint you can see — a different axis of vision that both Snellen and logMAR leave untouched.

Take the test now. It is free, runs in your browser in about three minutes, and gives you a baseline on the contrast axis you can retest over time. The calibration and adaptive-procedure details live on the methodology page. As always: a screening signal to bring to your eye doctor, never a substitute for the exam.

References

• Bailey, I. L., & Lovie, J. E. (1976). New design principles for visual acuity letter charts. American Journal of Optometry and Physiological Optics, 53(11), 740–745. The foundational paper establishing equal-letters-per-line, proportional spacing, and logarithmic size progression — the design principles behind the logMAR and ETDRS charts.
• Ferris, F. L., 3rd, Kassoff, A., Bresnick, G. H., & Bailey, I. (1982). New visual acuity charts for clinical research. American Journal of Ophthalmology, 94(1), 91–96. Introduced the ETDRS chart based on Bailey-Lovie principles; became the standard research acuity chart.
• Kaiser, P. K. (2009). Prospective evaluation of visual acuity assessment: a comparison of Snellen versus ETDRS charts in clinical practice (An AOS Thesis). Transactions of the American Ophthalmological Society, 107, 311–324. Found clinically meaningful discrepancies between routine Snellen and standardised ETDRS acuity in a real retinal practice.
• Pelli, D. G., Robson, J. G., & Wilkins, A. J. (1988). The design of a new letter chart for measuring contrast sensitivity. Clinical Vision Sciences, 2, 187–199. The Pelli-Robson chart — the contrast-domain analogue of the logMAR reform.
• Lesmes, L. A., Lu, Z.-L., Baek, J., & Albright, T. D. (2010). Bayesian adaptive estimation of the contrast sensitivity function: the quick CSF method. Journal of Vision, 10(3):17. The adaptive method behind modern efficient contrast-sensitivity estimation.