Why do sizes differ between brands, and what should you measure instead?

TL;DR

Sizes differ between brands because a size label names a product, not a person. ISO 8559-2 bases clothing size designation on body measurements and states that the garment's own measurements are the designer's and manufacturer's choice; ISO 9407 says the same of a shoe last. So the same US 9.5 foot is recommended a 27.5 boot by one brand chart, a 27.5 or a 26.5 by the second, and a 26.5 by the third, while the same 112 cm chest is an L in two outerwear charts and an XL in two others. The numbers that survive crossing brands are the ones taken on you: foot length, foot width, instep and chest girth, measured the way the standards specify.

The question, as people ask it

"I am a 42 in one brand and a 43 in another. Which am I?"

Neither. A 42 is not a measurement of you. It is the length of a wooden or plastic last, in a unit of two thirds of a centimetre, and the maker chose how much longer than a foot that last would be. You are 275 millimetres. That number is the same in every shop on earth, and it is the only one in this story that is.

Sizing explained: three kinds of number on a label

The European size standard EN 13402 opens by sorting every label into one of three kinds. A label can state the body it was made for. It can state the dimensions of the product. Or it can state "a size number or code with no obvious relationship to any measurement", which is what the standard calls an ad hoc size, and what the rest of us call a 12 or an XL.

Only the first kind names you. Here is where the labels you actually meet fall.

A shoe size is the second kind. In the Continental European system the size is the length of the last in Paris points of two thirds of a centimetre, and the last is typically 13 to 17 mm longer than the foot, or 2 to 2.5 Paris points, "so to determine the shoe size based on actual foot length one must add 2 Paris points". UK and US sizes are the same idea in barleycorns of a third of an inch, minus a constant, and the same reference notes that most US manufacturers use offsets larger than the nominal one, some by half a barleycorn and some by a whole one. Work your own case. A 275 mm foot is 27.5 cm, times three halves is 41.25, plus the two points of allowance is 43.25: a 43. A maker who allows the upper end of that range, two and a half points, lands at 43.75: a 44. The 42 and the 43 in your question are two makers' allowances. They are not two feet.

A letter size is the third kind dressed as the first. Each brand publishes the range of chests its M covers, and it looks like a body measurement. It is a body range the brand chose, and no standard fixes it. The five charts below show how far the choices drift.

A women's numeric size is the third kind with a history. The US commercial standard that once defined it, CS 215-58, was declared voluntary in 1970 and withdrawn entirely by the Department of Commerce in 1983, and the voluntary ASTM tables that followed have moved on their own. Read what a size 8 bust meant in each of the four published tables, before any brand had touched the label.

The drift did not stop at the standards. The vanity sizing record notes that in the 1937 Sears catalog a size 14 dress had a 32 inch bust; in 1967 that bust was a size 8; in 2011 it was a size 0. And a 2003 study that measured over 1,000 pairs of women's pants found that the more expensive brands ran smaller than the cheaper ones at the same nominal size. Since the withdrawal, in the words of the reference, "clothing brands and manufacturers size their products according to their preferences".

Why the same size decompresses differently

The mechanism is not mysterious, and it is not malice. It is three private choices per brand.

A brand fits one body first. Pattern grading is "the process of turning a sample size (sometimes referred as base size) into an additional smaller or larger sizes", and the sample size is one fit model's body. Then the brand adds its own ease: a shirt with a 42 inch garment chest is built to fit a 38 inch body, and the allowance runs from roughly 0 to 2 inches in a fitted cut to 5 or more in a relaxed one. Then it steps the pattern out by its own grade rule, where "in a standard size range this POM 'grades' 1-2 inches between sizes". Base body, ease, grade. None of the three is printed on the label, and every brand picks all three.

Now watch what that does to one foot. Ski boots are sized in mondopoint, the length of your foot in millimetres, so every boot chart has a mondo column that is the foot itself. The same charts also convert your shoe size to a mondo size, and to do that each brand has to guess how much longer than your foot your shoe's last was.

Three charts, one shoe size, a full centimetre of disagreement, and none of them is wrong. Sources: the Head boot chart, the Nordica sizing PDF and the Salomon boot chart, all read 2026-09-07. The brands know. Nordica's own sheet says: "Since shoe sizes vary across brands, it's best to measure the actual length of your foot in centimeters." Head's says: "Manufacturers do not always use the same sizes for shoes. We recommend using the foot length to find the correct size." Blister's bootfitting guide puts a number on the other route: conversion charts "would often recommend a Mondosize that's 1-2 sizes too big, if going off your preferred shoe size".

Now one chest. Take 112 cm, which is 44 inches, and ask five men's outerwear charts which letter it is.

Two charts say L, two say XL, one has no row for you. Same chest, same kind of jacket, same season, read from the brands' own published charts for The North Face, Patagonia, Burton, Salomon and Arc'teryx. To be fair to the letters, the same five charts cluster at the medium, all of them putting M somewhere around 99 to 104 cm. The labels do not diverge everywhere. They diverge unpredictably, and there is no way to know where without the number that does not move.

What to measure instead, and how

Four numbers survive crossing brands. Each is a measurement of you, each has a condition that makes it the standard's number rather than a guess, and each has its own piece on this site that goes deep.

The deep versions: how to find your mondopoint number, what last width is and how to measure yours, what instep height is and how to measure your chest and what the number leaves out.

Then one rule for every chart you meet. Read it from the body column inward, never from the letter outward. Find the column that is a measurement of a person, find your number in it, and only then read the letter or the size beside it. If the chart quotes garment measurements, subtract the ease for that cut before you compare. If the chart has letters and no measurements at all, it has told you nothing, and you should treat it that way.

What it trades against

Your measurements get you to the right row. They do not get you the fit, because the row's ease, its cut and its last are still the maker's, and that is the honest limit of this method. The label still has a job. It is the last step, not the first.

Two more things the measurement cannot do for you. It cannot make three sizing systems line up, because they were never built on the same grid.

Three rulers with three step sizes and three zero points never coincide, which is why the shoe size reference warns that conversion tables "often contain obvious errors, not taking into account different zero points or wiggle room", and why ISO's own conversion specification, ISO/TS 19407, describes itself as "a good compromise solution" and rounds every converted value up rather than down. A conversion is a guess about a last. A measurement is not.

And it cannot tell you how a given shell loads your foot once you are in it. A boot's flex rating changes what the same length and volume feel like across the instep, and the manufacturer number for it is not comparable between brands either.

The part a person cannot do

Every step above is easy. Four numbers, ten minutes, a sheet of paper and a tape.

The hard part is the other side of the table. Every candidate has a chart. Every chart is one brand's base body, its own ease, its own grade rule and its own last allowance, none of them printed, all of them different, and the only way to know where the labels diverge is to read every chart from the body column in. Holding your four numbers against all of that, for every boot and every jacket, is not a research problem. It is a working memory problem, and the people who are good at it do it for a living with the product in their hands.

You should not have to be that person. State your numbers once and let software carry them across every candidate.

openclaw plugins install clawhub:@4gpts/sil

Then say: My foot is 275 mm long and 102 mm wide standing, my instep is 2 cm over my foot length, my chest is 104 cm, and I ski advanced. Find me boots and a shell jacket that fit, and tell me why.

It reasons over your length, your width, your instep and your chest, against each chart's body column rather than its letters. The labels stay exactly as compressed as they were. What changes is that they stop being the only thing you were given to think with.

FAQ