AbilityBenchFree, 2 methods × 4 trials, nothing to calibrate
Dominant eye test that runs both methods, several times each
Point at something across the room, close one eye and then the other, and one of them keeps your finger on target while the other makes it jump — that eye is your sighting-dominant one. This page runs that test and the older hand-aperture version four times apiece, tells you which eye to close first on each trial so the instruction cannot bias the answer, and reports the proportion of trials that agreed rather than a verdict from one attempt. It is free, needs no account, and calibrates nothing, because the whole test happens in the room rather than on the screen.
- 100% free
- No signup
- 8 trials by default
- 2 methods
- No screen setup
One eye is doing the aiming. When you sight along a rifle, look down a microscope, hold a camera to your face or line a picture hook up with the wall, your brain picks one eye and uses it — and most people have never checked which. This page runs the two oldest sighting tests 4 times each, in three gaze directions, and reports how often the same eye won rather than announcing an answer from one attempt.
Before you start
- Find a small fixed object at least 3 m away that you can see from your chair — a door handle, a light switch, a corner of a window frame. Closer than that and your eyes converge enough that the aperture sits over both of them.
- You will be asked to turn toward targets to your left and right as well as straight ahead, so pick a spot with something to look at in all three directions, or be ready to swivel.
- Answer with the L and R keys, or N when neither eye held the object. Backspace throws the last answer away and asks it again, which is what to use when you realize you closed the wrong eye.
When this measurement is not the one you want
Both tests need two eyes that are pointing at the same thing and two hands you can hold in front of your face. If one eye is patched, amblyopic, or turned by a squint, the answer will be the other eye every time and it is describing that, not a preference. If you cannot raise both arms, run the pointing method alone with one finger and read the Porta row on its own — that is a real halving of the evidence rather than a workaround.
How to find your dominant eye properly
Pick a far target, follow the trial instructions, and read the split rather than the verdict.
Pick something at least 3 meters away
A door handle, a light switch, the corner of a window frame — small, fixed and well across the room. Distance matters more than people expect: closer than about three meters your eyes are converging noticeably, the near aperture stops being aligned with just one of them, and the test starts answering a question about convergence instead. You will also be asked to look well to your left and well to your right on some trials, so choose a spot where there is something to aim at in three directions.
Do exactly what the trial says, including which eye closes first
Each trial names the method, the direction to look in, and which eye to shut before the other. That last instruction is not decoration. On the pointing test the finger jumps the moment the aligned eye closes, and it is easy to report whichever eye you happened to close second simply because that is the view you were left holding — so the page alternates the order and shows you afterwards whether your answers split by it.
Answer with L, R or N, and use Backspace when you fumble one
Press L or R for the eye that held the target, or N when it stayed put with both or moved with both. Backspace throws away the trial you just answered and asks it again, which is what to reach for when you realize you closed the wrong eye or lost the target — the discarded trial is counted separately and is not in the totals. There is no clock on any of this, so take as long as raising two arms takes.
Technical specifications
| Methods run | Two, in blocks: the Miles hand-aperture test, in which both eyes stay open while a triangular hole between your palms is aimed at the target, and the Porta test, a single pointing finger with no aperture at all |
|---|---|
| Trials | 4 per method by default, selectable as 2, 4 or 6 — so 4, 8 or 12 in total. Six per method is the smallest run in which a 5-to-1 split says much, and the page shows why by printing the interval that comes with your denominator |
| Gaze directions | 3 — well left, straight ahead, well right — cycled from a shuffled triple inside each block rather than drawn independently, because an independent draw of four trials leaves one direction missing about a third of the time |
| Order control | The eye you are told to close first alternates trial by trial, and the results panel splits your answers by that order so an instruction effect shows up as a column rather than hiding inside the total |
| Answer set | Left, right, or undecided. Undecided is a real outcome kept out of the numerator and the denominator both, so a run where the target barely moved reports a small denominator instead of a coin-flip verdict |
| What is computed | The count for the winning eye over the trials that named one, a 95% Wilson score interval on that proportion, the per-method split, the per-direction split and whether the two methods agreed |
| Minimum target distance | 3 m, stated because it is the one physical condition the page cannot check and the commonest reason two methods disagree on a person whose eyes are otherwise settled |
| The reference beside your split | ≈ 65% of adults are right-eyed on a sighting test — Bourassa, McManus & Bryden (1996), Handedness and eye-dominance: a meta-analysis of their relationship, Laterality. It is a prevalence share, not a distribution, so it tells you which group you are in and supports no percentile |
Frequently asked questions
Miles and Porta gave me different eyes. Which one is right?
Both are, and the disagreement is the finding. The two tests differ in how far the near object sits from your face and in whether both eyes are open while you aim, so they do not sample the same alignment — and a person whose two eyes are close in strength can land on either side of that difference. The useful reading is not to pick a winner but to treat the disagreement as weak dominance, which is a genuine and common state. If you need a single answer for a practical purpose, the pointing test is the one that resembles most real aiming tasks, because most of them involve one hand and no aperture.
Is the dominant eye the stronger eye, or the one with better vision?
No, and conflating them is the most common error in this corner of the internet. Sighting dominance is about which eye an alignment task defers to; acuity is about how fine a detail an eye resolves. They are measured differently, they are not the same trait, and plenty of people sight with the eye that has the worse prescription. The literature is blunter still: Mapp, Ono and Barbeito (2003), What does the dominant eye dominate? A brief and somewhat contentious review, Perception & Psychophysics, went looking for functions in which the sighting eye holds a privileged position and found the case much weaker than the popular account assumes.
Does my dominant eye have anything to do with being right-handed?
The link is real but far weaker than people expect, which is why a left-eyed right-hander is ordinary rather than curious. ≈ 65% of adults are right-eyed against a much larger share who are right-handed, so the two cannot be tracking each other closely — Bourassa, McManus & Bryden (1996), Handedness and eye-dominance: a meta-analysis of their relationship, Laterality pooled the studies and found a modest association rather than a rule. Being right-handed is a mild indication of being right-eyed and nothing you should plan a rifle stock around.
Why does the page make me look sideways as well as straight ahead?
Because for many people the answer changes when they do. Khan and Crawford (2001), Ocular dominance reverses as a function of horizontal gaze angle, Vision Research, had observers sight at targets across a wide range of horizontal eccentricities and found the sighting eye switching in most of them once the target moved far enough to the side. A test run once, straight ahead, cannot see any of that. Three directions is the least this page can do about it, and the per-direction table is where a reversal shows up.
Can I do this on a phone, or with the screen at all?
You can read the instructions on a phone, but nothing here is drawn on the screen and it could not be: a sighting test needs a near object and a far one so the two eyes disagree about their alignment, and a flat panel has exactly one distance. Any site that shows you two shapes and claims to have found your dominant eye from what you tapped has measured something else — usually which side of a pair you preferred. That is also why this is the only page in this group that asks for no card and no viewing distance.
Which eye should I use for a camera, a microscope or a rifle?
Start with the one this page reports and then let the equipment argue back. Rifle and bow shooting genuinely want the sighting eye behind the sights, and a mismatch with your handedness is worth knowing before you buy anything, which is the practical reason this test has survived four centuries. Microscopes and camera viewfinders are looser — plenty of photographers use the non-dominant eye deliberately, to keep the other one open on the scene — and a monocular sight you can hold steady beats the eye a test picked out. Where the two eyes have to cooperate rather than take turns, the relevant measurement is not dominance at all.
I got 4 right and 4 left. Does that mean the test failed?
No, it means you do not have a strong sighting preference, and that is a result rather than a null. The phrase dominant eye implies a binary that a fair number of people simply do not have; run for run they alternate, and the split is what an honest test reports. Two things are worth checking before you accept it: that your target really was three meters away or more, and that your answers did not split by the eye you were told to close first, which the results panel breaks out for exactly this reason.
Two four-hundred-year-old tests, and what dominance does not mean
The pointing version is the older of the two by a long way: Giambattista della Porta described it in 1593, and it has not needed a revision since, because the geometry is simply that a finger held between your face and a distant object is aligned with it for one eye and displaced for the other. The aperture version comes from Miles (1929), Ocular dominance demonstrated by unconscious sighting, Journal of Experimental Psychology, and its improvement is in the word unconscious: both eyes stay open while the hole is aimed, so the visitor never chooses an eye, and the sighting one is revealed only afterwards when the other is closed. That is a better piece of experimental design than the pointing test, and it is still not enough on its own, which is why this page runs both and reports whether they agreed.
What almost every online version gets wrong is the sample size, not the procedure. One trial produces one of two answers, so it cannot distinguish a person who sights with the right eye every time from a person who does it slightly more often than not — and the second group is large. Repeating the trial makes the distinction available, and the price is that the page then has to be honest about a small denominator: eight trials that all agree still leave a 95% Wilson interval reaching down into the sixties, and the panel prints that interval rather than rounding it away. The other thing repetition buys is a look at the two known instabilities — the disagreement between methods, and the reversal with gaze angle that Khan and Crawford measured — both of which are invisible to a single attempt and both of which get their own row here.
The last thing worth saying is what dominance is not, because the popular account overreaches in a specific direction. A sighting-dominant eye is not the eye with better acuity, not the eye that leads in every visual function, and not a fact about how well your two eyes work together. That last question is a separate measurement, and it is the one people are usually after when they arrive here worried: whether the eyes combine at all is what the stereopsis test asks with random-dot depth, and how well you judge distance from the cues that survive on a flat panel is what the depth perception test reports. If what brought you here was curiosity about how much the two eyes differ, the single-eye measurements are the informative ones — cover one at a time in the blind spot test and compare the two maps, or run the peripheral vision test once per eye and compare the fall-off. Both of those need a card and a viewing distance; this page needed nothing but a wall three meters away.
This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish a squint, amblyopia, or any other reason your two eyes might not point at the same thing. Only a qualified professional, working with more than a browser, can make that judgment.
Where your eight answers go
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Your answers, your reaction times and your score are never uploaded, logged or kept — which is also why the test carries on working after you disconnect from the network, and why nothing here can be held back behind an email address.
The run is an array of eight small records in this tab — method, direction, close-first eye, the letter you pressed — scored in place and gone when you close the page. Nothing is written to local storage and no seed is put in the address bar, so a reload starts a fresh run; the copy button is what gets a result out, as plain text carrying the split, the interval, both method rows and the run seed.