Ligne: The Unit Buttons Are Sold In
A button size chart is only useful once the unit makes sense, and the button trade uses one that appears almost nowhere else. Buttons are sized in ligne — written L, sometimes spelled "line" — and one ligne is exactly one fortieth of an inch. That fixes everything else: 1 L = 0.635 mm, and 40 L = 1 inch. A 16 L button is 16/40 of an inch, or 10.16 mm.
The arithmetic runs both ways in one step. Multiply ligne by 0.635 for millimetres; divide millimetres by 0.635 for ligne. A 24 L button is 15.24 mm. A button you have measured at 20 mm is 31.5 L, which rounds to the 32 L that is actually made.
The other ligne, and why it matters
There is a second unit with the same name, and mixing them up produces answers that are wrong by a factor of three and a half. The historical French ligne is 2.2558 mm — one twelfth of a Paris inch, a pre-metric unit that survives today mainly for sizing watch movements. The button ligne is 0.635 mm. They are unrelated units that happen to share a word. If a conversion result looks absurd, this is usually why.
Button Size Chart: Ligne to mm to Inches
These are the sizes actually manufactured. The series has gaps on purpose — buttons come in a conventional set of lignes, so there is no 33 L to buy even though the arithmetic would happily give you one.
| Ligne | Millimetres | Inches | Sold as |
|---|---|---|---|
| 12 L | 7.62 mm | 0.300 in | — |
| 14 L | 8.89 mm | 0.350 in | 9 mm |
| 16 L | 10.16 mm | 0.400 in | 10 mm |
| 18 L | 11.43 mm | 0.450 in | 11 mm |
| 20 L | 12.70 mm | 0.500 in | 1/2 in |
| 22 L | 13.97 mm | 0.550 in | 9/16 in |
| 24 L | 15.24 mm | 0.600 in | 5/8 in |
| 27 L | 17.15 mm | 0.675 in | 11/16 in |
| 28 L | 17.78 mm | 0.700 in | — |
| 30 L | 19.05 mm | 0.750 in | 3/4 in |
| 32 L | 20.32 mm | 0.800 in | 13/16 in |
| 34 L | 21.59 mm | 0.850 in | 7/8 in |
| 36 L | 22.86 mm | 0.900 in | 15/16 in |
| 40 L | 25.40 mm | 1.000 in | 1 in |
| 45 L | 28.58 mm | 1.125 in | 1 1/8 in |
| 50 L | 31.75 mm | 1.250 in | 1 1/4 in |
| 54 L | 34.29 mm | 1.350 in | — |
| 60 L | 38.10 mm | 1.500 in | 1 1/2 in |
The millimetre and inch columns are computed from the definition rather than copied from a supplier's chart, and the "sold as" column shows why that distinction is worth making. Trade labels are rounded to names a merchant can print: 36 L is sold as "15/16 inch" though 36/40 is exactly 0.9 inch, and 15/16 is 0.9375. The labels run a little generous. The ligne number is the exact one — prefer it whenever both are given.
Converting mm to line
"Line" is simply the anglicised spelling of ligne, so a button size chart in mm to line is the same conversion under another name, and a "24 line" button is a 24 L button. The ligne-to-mm conversion is one multiplication in either direction — ligne × 0.635 for millimetres, millimetres ÷ 0.635 for ligne — which is why the table below carries both columns rather than making you do it.
A Printable Button Size Chart That Cannot Mislead You
There is a version of the printable button size chart that is perfectly safe, and the distinction is worth being precise about, because this page has just spent a section warning about printed charts.
Numbers are scale-independent; pictures are not. If a printer shrinks a sheet by four percent, "24 L = 15.24 mm" is still true — the text is smaller, the fact is unchanged. It is only a true-size circle that a scaling error silently corrupts, because its whole job is to be a specific physical width. So print the table, and size the button on the screen where calibration is visible and checkable.
That gives you a button size chart PDF with the navigation stripped out — no email, no sign-up, no download wall — using your browser's own "Save as PDF", which exists on every desktop OS and on iOS and Android. Keep it for the conversions. Do not measure against it.
What Size Button Does a Garment Take?
Half the reason people look up a button size chart is not to identify a button but to choose one. These are the conventional ranges, and the gauge above will highlight any of them.
| Use | Ligne | Millimetres |
|---|---|---|
| Shirt collar / button-down | 13–14 L | 8.3–8.9 mm |
| Shirt sleeve placket | 15–16 L | 9.5–10.2 mm |
| Shirt front | 17–18 L | 10.8–11.4 mm |
| Sport / casual shirt | 19–22 L | 12.1–14.0 mm |
| Trousers | 23–27 L | 14.6–17.1 mm |
| Suit jacket sleeve | 23–24 L | 14.6–15.2 mm |
| Suit jacket front | 29–36 L | 18.4–22.9 mm |
| Overcoat | 40–45 L | 25.4–28.6 mm |
Treat these as conventions, not specifications. Published charts genuinely disagree, and the jacket row is the worst of it: WAWAK's chart puts suit fronts across 29–36 L, while other trade charts begin jacket fronts at 36 L and run to 40. The ranges above are drawn wide enough to span that disagreement rather than pick a winner. Shirts are the exception — charts broadly converge on 17–18 L for the front, a size down for the sleeve, and smaller again for the collar.
For a repair, none of this beats measuring a surviving original. A garment's own buttons are ground truth about what that garment takes; a chart is an average over a trade.
Measuring a Button You Already Own
Button size means the diameter of the face, measured straight across through the centre at the widest point. For square, oval and novelty buttons, trade practice is to take the longest dimension across the face.
Two things are commonly measured by mistake. The shank does not count — the loop or stem on the back of a shank button is not part of its stated size, so lay such a button face-down on the gauge with the shank pointing up and the face flat on the glass. And on a domed or rimmed button, the widest point may not be the visible top; it is often a rim lower down, which is what a ring will catch and a quick eyeball will not.
Two specifications the ligne number says nothing about: shank height and hole diameter. A replacement matched on ligne alone can still sit at the wrong height on thick cloth, or refuse the thread weight you intended. Neither appears on any button size chart, this one included, so check them against the original when the fit matters.
If you have digital calipers, use them and divide by 0.635. Most people do not, and this is the case where a ruler struggles: the sizes step by 2 L, which is 1.27 mm, and reading a curved edge against ruler graduations at that resolution invites parallax error in exactly the direction that loses you a size. Comparing against a true-size ring sidesteps the reading entirely — either the outline disappears behind the button or it does not.
Why "Just Print the Chart" Goes Wrong
Search for how to size a button and the advice converges on printing a chart and laying the button on it. One of the sites giving that advice sells the printable. The method is sound in principle and fails in practice for a mundane reason: almost nothing prints at exactly 100% by default. Browsers and print dialogs apply "fit to page" or shrink-to-fit-margins scaling, which silently rescales every circle on the sheet. A few percent small is enough to read every button one size low, and nothing on the printed page reveals that it happened.
A printed chart is only trustworthy if you check the printed sheet against a ruler afterwards — at which point the ruler could have measured the button directly. Calibrating a screen has the same requirement, with one difference: it is done once, visibly, against an object of known size, and it is obvious when it is wrong.
That calibration carries across this site. The value is shared with the online ruler, crochet hook size chart, knitting needle gauge and fabric count gauge, so calibrating on any one of them calibrates the rest on that screen.
How much to trust it
Better than most uses of this technique, because the target is forgiving. Manufactured sizes sit 2 L apart — 1.27 mm — and a careful bank-card calibration lands comfortably inside that. The failure mode to watch is browser zoom: calibrate at 100%, and recalibrate if you change it, because zoom rescales the rings and the card outline together and the outline will still look correct. Where a reading falls between two rings, the chart above gives the exact millimetre figures for both.