Quick answer: near PD is not a fixed subtraction from distance PD. It is a geometric consequence of where the spectacle plane sits relative to the eyes’ centres of rotation, and the standard formula is near PD = distance PD × l / (l + s), where l is the working distance and s is the distance from the spectacle plane to the centre of rotation, conventionally taken as 27 mm. At a 40 cm working distance that produces a factor of 0.937, which is where the familiar “subtract 4 mm” rule comes from. At 25 cm the correct total inset is about 6.2 mm, so the shortcut is off by more than 2 mm. The calculator below runs the real geometry per eye. It does not measure a PD: bring a measured distance PD with you.
Where the Formula Comes From
When the eyes converge on a near object, the visual axes cross the spectacle plane closer together than the pupils themselves are. What matters optically is not the separation of the converging pupils but the horizontal separation of the converging visual axes where they intersect the lens. That distance is the near centration distance, or near PD.
The geometry is a pair of similar triangles. The eyes rotate about their centres of rotation, which sit behind the spectacle plane. Following Mo Jalie’s treatment in The Fitting of Spectacle Lenses:
monocular near PD = monocular distance PD × l / (s + l)where l is the working distance and s is the distance from the spectacle plane to the eye’s centre of rotation. In practice s is assumed to be 27 mm, which bundles a typical vertex distance together with the position of the centre of rotation behind the cornea.
Because s is a constant, the whole expression collapses into a single multiplier for any given working distance:
| Working distance | Factor applied to distance PD |
|---|---|
| 25 cm | 0.903 |
| 33.3 cm | 0.925 |
| 35 cm | 0.928 |
| 40 cm | 0.937 |
Those four factors are published in the same source, which makes them a useful check on any implementation, including ours.
Why the “Subtract 3 to 5 mm” Rule Drifts
The traditional shortcut is to take 3 to 5 mm off the binocular distance PD, or 1.5 to 2.5 mm off each monocular value. It survives because it is very nearly right for an average PD at a 40 cm working distance, which describes most reading prescriptions.
It drifts in two directions at once, because the formula is multiplicative and the rule is additive.
| Distance PD | Working distance | Calculated near PD | “Subtract 4 mm” | Error |
|---|---|---|---|---|
| 64 mm | 40 cm | 60.0 mm | 60.0 mm | 0.0 mm |
| 72 mm | 40 cm | 67.4 mm | 68.0 mm | 0.6 mm |
| 54 mm | 40 cm | 50.6 mm | 50.0 mm | 0.6 mm |
| 64 mm | 25 cm | 57.8 mm | 60.0 mm | 2.2 mm |
The pattern is worth internalising: the working distance moves the answer far more than the PD does. A 64 mm PD held at 25 cm needs 6.2 mm of total inset, not 4 mm. That is the case where the shortcut is genuinely unsafe, and it is not exotic: it covers close handwork, jewellers, dentists, musicians reading from a stand set close, and anyone whose near task is not a book at arm’s length.
Whether a 2 mm centration error matters depends on the power in the lens. Through a +2.00 D lens a 2 mm error induces 0.4 prism dioptres of unwanted prism by Prentice’s rule, which is tolerable for most wearers. Through a +6.00 D post-cataract reading lens the same error induces 1.2 prism dioptres, which is not.
When You Actually Need a Near PD
This is the part most calculators leave out, and getting it wrong wastes effort or causes a remake.
You need a near PD for:
- Single vision reading glasses, where the optical centres are placed at the near centration distance
- Occupational and near-variable-focus lenses ordered to a specified working distance
- Any lens where the lab asks for a near centration distance explicitly
You do not order a near PD for standard progressives or bifocals. Those designs have the inset built into the geometry by the manufacturer, computed from the add power and the design’s own assumptions about working distance. Sending a near PD where the lab expects a distance PD moves the distance zone inward by the inset amount and produces a lens that is wrong at every distance. The value the lab wants for a progressive is the monocular distance PD plus the fitting height.
The exception is a personalised progressive whose ordering form asks for the wearer’s actual working distance. There the number feeds the design calculation rather than replacing the distance PD, and the calculator on this page is not what generates it.
Measure Monocularly, Because Inset Is Per Eye
The formula operates on monocular values for a reason. Each eye converges independently toward the near point, so each eye has its own inset, and a face whose distance PD splits 33 / 31 does not become symmetric at near.
Halving a binocular PD and applying the factor gives the correct binocular total but the wrong per-eye split, which is exactly the number the lab uses to place each optical centre. The calculator accepts a binocular entry for convenience and tells you it is assuming symmetry, but monocular input is the mode that produces a usable answer. The distinction between the two, and when each is appropriate, is covered in our monocular vs binocular PD guide.
What This Calculation Assumes
Three assumptions are baked in, and all three are standard rather than sloppy. They are worth knowing because each one has a case where it breaks.
- A flat spectacle plane. Most near PD calculations assume a 0 degree wrap angle. A strongly wrapped frame moves the lens plane away from that assumption, and the geometry no longer holds exactly.
- A 27 mm centre-of-rotation distance. This is a population convention, not a measurement of the wearer in front of you. The centre of rotation shifts with ametropia, and vertex distance varies with how the frame sits. The calculator exposes
sas an advanced input for anyone who wants to adjust it deliberately. - No prismatic compensation. A significant distance prescription bends the ray path before it reaches the near point, so the true near centration in a high-powered lens differs from the pure geometric result. Treat the output as the geometric baseline it is.
None of these invalidate the calculation. They just mean the result is a well-founded starting point rather than a measured fact, and that the input matters more than the arithmetic: a distance PD that is 1 mm off produces a near PD that is about 0.94 mm off, no matter how precise the formula. Accuracy requirements for centration are covered in our guide to prescription and lens tolerances.
Frequently Asked Questions
What is the formula for near PD?
Near PD equals the distance PD multiplied by the working distance, divided by the working distance plus 27 mm, with both distances in the same units. The 27 mm represents the distance from the spectacle plane to the eye’s centre of rotation. Applied per eye, the same formula gives each eye’s near centration distance and inset.
How much smaller is near PD than distance PD?
At a 40 cm working distance, about 4 mm smaller for an average adult PD, which is where the traditional rule of thumb comes from. The difference grows as the working distance shortens: at 25 cm a 64 mm distance PD gives a near PD of about 57.8 mm, a difference of 6.2 mm.
Do I need a near PD for progressive lenses?
No. Standard progressive and bifocal designs have the near inset built in by the manufacturer, calculated from the add power. Ordering a progressive with a near PD in place of the distance PD decentres the distance zone and produces a lens that is wrong at all distances. Progressives need a monocular distance PD and a fitting height.
Why is 27 mm used in the formula?
It is the conventional value for the distance between the spectacle plane and the eye’s centre of rotation, combining a typical vertex distance with the position of the centre of rotation behind the cornea. It is a population convention rather than a per-patient measurement, which is why the calculator lets you override it.
Can I just subtract 4 mm from the distance PD?
For an average PD at a 40 cm working distance, that shortcut lands within a few tenths of a millimetre. It degrades as the PD moves away from average and, more sharply, as the working distance shortens. At 25 cm it under-corrects by more than 2 mm, which matters in higher-powered reading lenses.
Does this tool measure my PD?
No. It converts a distance PD you already have into a near PD. You need to supply the measured distance PD as an input, taken with a pupillometer, a corneal reflection method, or a digital measurement system.
Sources
- Mo Jalie, The Fitting of Spectacle Lenses (via ABDO) (near centration distance formula, the 27 mm centre-of-rotation assumption, the published working-distance factors and the reference table used to validate this calculator)
- Meet Your Pupilometer, 20/20 Magazine (the variables affecting near PD, including vertex distance, centre-of-rotation position and the flat spectacle plane assumption)
I grew up inside an optical shop. My mother likes to tell how, as a kid, I would watch eyeglasses being assembled, play with the tools, and draw on the back of service order slips. The family business taught me early what a precise measurement and work done right are worth.
I went on to spend more than twenty years as a software engineer, always keeping an eye on the optical world. Optogrid was born where the two meet: digital measurement technology (pupillary distance and fitting heights) built by someone who knows the dispensing counter from the inside.