How to Measure PD With a Ruler: The 18-Inch Setup and the Number It Misses By

One number or two: what a PD is

Pupillary distance is the distance between the centres of the two pupils, in millimetres. It is written in two forms, and the difference matters more than the names suggest. A single PD is one number: the distance from one pupil centre to the other, taken as a straight line across the face. A dual PD, also called monocular PD, is two numbers: the distance from each pupil centre to the midline of the bridge of the nose, right eye first. Cleveland Clinic's example is 31/32, which adds to 63 — the same face, written two ways.

The single number is enough for a pair of single-vision glasses made to a symmetrical frame. It is not enough when the two eyes sit at different distances from the midline, which is common and which a single number averages away. Dual PD is what a progressive or a high-power lens needs, because the two lenses are not interchangeable and each one has to be centred on its own eye. The arithmetic that uses those numbers lives on the frame and lens pages this one links to; the job here is only to get them.

One caution before the ruler comes out. PD is not a constant for life. It grows through childhood and then usually holds once the face is done developing, but it can change by a small amount between distance and near focus, and it can be affected by conditions that change eye alignment. That is why the number on an old prescription is a starting point rather than a permanent fact.

How to measure your PD at home, alone

Stand in front of a mirror in a well-lit room with your face about 18 inches (46 centimetres) from the glass. Hold a ruler marked in millimetres against your face, just under the eyes, with the markings facing the mirror so you can read them in the reflection. Hold it the way you would hold a fence you were peeking over: steady against the cheekbones, not pressed into the eyes.

Close your left eye. With your right eye, line the zero mark of the ruler up with the centre of your right pupil — the imaginary vertical line that divides the black centre into a left half and a right half. Do not move the ruler after this. Open your left eye, close your right, and find the nearest millimetre mark to the centre of your left pupil. That mark is your single PD.

The same setup gives you dual PD with one extra step. Keep the ruler where it is, and read the distance from the centre of each pupil to the midline of the bridge of the nose. The two numbers should add to the single PD you just read. If they do not, one of the three marks moved; reset and take all three again rather than trusting the pair that agrees.

With a helper the procedure changes at the pupils, and the change is worth knowing. Cleveland Clinic's helper version lines the zero mark with the outer edge of the right pupil and reads to the inner edge of the left pupil, while the solo version uses the centres of both. Those are different spans, not two ways of saying the same thing. Pick one version and use it for both attempts. Mixing them produces a difference that looks like measurement error and is actually a change of definition.

Two attempts, ten minutes apart, are the minimum this page asks for. If they agree to the millimetre, write the number down. If they differ by one millimetre, take a third and use the two that agree. If they differ by more than that, the ruler moved, the head moved, or the pupils were not where you thought; start again rather than averaging the disagreement.

The average, the range, and why two published ranges differ

Cleveland Clinic puts the average adult single PD at about 63 mm and the usual range at 50 to 70 mm, with values outside that range described as very rare. Dodgson's 2004 review, built on the ANSUR database of 3,976 adults, gives a mean of 63.36 mm and a standard deviation of 3.83 mm, with the sample running from 52 to 78 mm. The same review says the vast majority of adults fall between 50 and 75 mm, and that 45 to 80 mm is likely to include almost every adult.

Those two ranges are not in conflict; they are answering different questions. Cleveland's 50 to 70 mm is a clinical description of what is usual. Dodgson's 50 to 75 mm is a description of where the vast majority of a measured sample sits, and his wider 45 to 80 mm is a design range meant to include nearly everyone. A person at 72 mm is outside Cleveland's usual range and inside Dodgson's vast majority at the same time. Neither source is wrong, and neither number should be read as a target.

The distributions also differ by sex and by viewing distance, which is where a single average loses information. In the ANSUR sample the male mean was 64.67 mm and the female mean 62.31 mm, a difference of about 2.4 mm. A 2021 study of 199 people found male binocular PD about 2 mm wider than female, and distance PD about 3 mm wider than near PD. A 2024 study of 40 people put the distance means between 63.6 and 64.8 mm across four instruments, and the near means between 61.1 and 62.1 mm. The near reading is smaller, consistently, by roughly 3 mm on average.

The inch column is the millimetre figure divided by 25.4, exact since 1959, and rounded separately — which is why a converted figure can differ from the printed inch by a hundredth. The standard deviations are measures of how far the people in that sample typically sat from its mean, not tolerances for an individual measurement. Cleveland Clinic's 50 to 70 mm and Dodgson's 50 to 75 mm are not competing estimates of the same quantity: the first is a clinical description of the usual range, the second a statement about where the vast majority of a measured sample fell. The 2024 study's instruments were a pupillometer, a PD ruler, an autorefractor and a phone application; its near readings are the ones to use when the question is about reading distance, and its distance readings are the ones to compare with a prescription written for distance.

What the ruler gets wrong, in millimetres and in prism

The 2024 comparison is the most useful published check on the ruler because it measured the same 40 people four ways and reported the differences rather than only the averages. Against the pupillometer, the ruler read 0.54 mm wider at distance, with a standard deviation of 0.74 mm and limits of agreement from −0.91 to 1.98 mm. At near it read 1.08 mm wider, with a standard deviation of 0.99 mm. The averages are small; the limits of agreement say that an individual reading can sit about a millimetre on the other side of the reference, and occasionally further.

Millimetres become more or less important depending on the lens. The Prentice rule converts a decentration into prism: prism dioptres equal the decentration in centimetres multiplied by the lens power in dioptres. Apply the study's average differences and the arithmetic runs like this. At 2.00 D, 0.54 mm is 0.11Δ and 1.08 mm is 0.22Δ. At 3.00 D they are 0.16Δ and 0.32Δ. At 4.00 D they are 0.22Δ and 0.43Δ. At 6.00 D they are 0.32Δ and 0.65Δ. The study's own conclusion was that every method's induced prism stayed inside the 0.33Δ tolerance it cites from ISO 16034:2002 — in that sample. The arithmetic here shows where that margin runs out: at 6.00 D the average near overread alone is about twice the tolerance, and the distance overread is right at it.

This is not an argument against the ruler. It is an argument for knowing which number you are holding and what power it is going into. A ruler PD on a −1.00 D pair is a different object from a ruler PD on a −6.00 D pair, even though the millimetre figure is identical. The first has room for the error; the second does not.

The prism figures are the mean differences from Jung and Chu (0.54 mm at distance, 1.08 mm at near) converted to centimetres and multiplied by the lens power. The 0.33Δ figure is the tolerance the same paper cites from ISO 16034:2002; it is a reference line, not a claim that a reading above it is dangerous. A real lens also has a prescription-specific centration, a frame fit and a vertex distance, none of which this four-row table includes. The table exists to show the scale at which a sub-millimetre difference starts to matter, not to grade any particular pair of glasses.

Near, distance, and the number a progressive needs

A single PD taken while looking at a mirror is a distance PD. Reading glasses and the near portion of a progressive are centred for a closer viewing distance, and the eyes converge slightly when they do. The 2024 study found near PD about 3.1 mm smaller than distance PD with the pupillometer, and about 2.7 mm smaller with the ruler. The 2021 study found the same direction, at about 3 mm. That is a sample average, not a fixed subtraction: do not take 3 mm off your distance PD and call it your near PD.

For a progressive lens the relevant numbers are the dual PD and the fitting height, and the second one is a property of the frame rather than of the face. Cleveland Clinic lists progressive lenses among the cases where a professional measurement is worth having, alongside bifocals, trifocals, and any pair that has to serve both close reading and distance vision. The reason is not that a ruler cannot produce a number. It is that the number has to be tied to a particular lens in a particular frame, and the frame is not on your face when you are standing at the mirror.

If you are ordering single-vision glasses online and your prescription is modest, a careful ruler reading taken twice is a reasonable input. If you are ordering progressives, high-index lenses, or anything with prism, take the ruler reading as a check on the professional measurement rather than as a replacement for it.

When to stop measuring and book the test

Cleveland Clinic names three situations that call for a professional measurement: special lenses such as bifocals, trifocals or progressives; a pair that has to work for both close reading and distance; and any condition such as strabismus that can affect eye alignment. To those, add two practical triggers of this page's own, and label them as such. First, if two careful attempts differ by more than 1 mm, stop and have it measured. Second, if your prescription is above about 4.00 D in either eye, the Prentice arithmetic above is the reason to have the number taken by someone with a pupillometer rather than a ruler.

The professional measurement is not a rebuke of the home one. A PD ruler and a pupillometer are measuring the same anatomy with different tools, and the published comparison puts their average difference below a millimetre at distance. What the professional adds is a second pair of hands, a head that is not moving, and a device that does not depend on reading a reflection in a mirror.

The record, in the order it arrived

Four sources carry every number on this page, and they are worth seeing in order because the later ones are checks on the earlier ones rather than replacements for them. The 2004 review established the population figures; the 2021 and 2024 studies measured how much the instrument changes the answer; the 2026 clinical page states the procedure in the form most people will actually follow.

Identities are given by kind of source and by date, not by link. The 2026 clinical page is the only row that is a patient-facing instruction; the other three are measurement studies or a review, and none of them was written to teach a home procedure. The 2004 review's ranges are population descriptions, not clinical cutoffs. The 2021 and 2024 studies used young adult samples, with mean ages of 24.1 and 25.65, so their averages should not be read as norms for children or for older adults.

Questions this page answers