why am i a medium here and a large there.

Published August 22, 2026·File 12 · Sizing·11 min read

Nothing forces a garment labelled M to measure anything in particular. We pulled eleven published size charts and spec sheets and put every medium side by side — including our own, which comes out of it looking small.

There is a drawer. Everyone reading this has the drawer. It contains four things labelled medium, and they are not the same garment. One fits like it was cut for you. One is a crop top now. One you have never worn in public and never will. One you keep because it was expensive and admitting the mistake costs more than the shelf space.

The reasonable conclusion is that your body is doing something strange between purchases. It is not. The letter on the label is not a measurement, is not regulated, and is not promising you anything. Once you know what it actually is, the drawer stops being a mystery and starts being a predictable outcome.

Why the letter is not a measurement

Start with what actually governs apparel sizing, because most people assume something does. Three documents matter. ASTM D5585-21 publishes standard tables of body measurements for one figure type across sizes 00 to 20. ASTM D6240/D6240M-24a does the equivalent for adult chests from 34 to 52 inches. ISO 8559-1:2017 defines how to take an anthropometric measurement in the first place — where the tape sits, how the subject stands.

Read that list again and notice what is missing. Every one of them describes a body. None of them describes a garment. They are reference tables for people who draft patterns, not rules for people who print labels. They are also voluntary — ISO 8559-1 is explicitly written as a guide for practitioners creating size and shape profiles, not as an enforceable specification. Nobody is audited against them.

So the chain from your chest to the letter on your label has a break in it. A brand chooses a fit model — one real person whose body the base size is drafted on. It chooses how much ease to add, meaning the gap between the body and the finished garment, which is a pure design choice and can be negative on a rib knit. Then it grades up and down from that base and assigns letters. Two brands can run that process honestly, competently, and to standard, and arrive at mediums that are five inches apart. Neither did anything wrong. There was never a rule to break.

So we measured it

The pages that rank for this question tend to stop at "there is no universal standard" and leave. That is true and it is not useful, because it does not tell you how much drift to expect or where. So on 22 August 2026 we pulled every size chart and manufacturer spec sheet we could source directly from the brand — no aggregators, no resellers — and put medium side by side.

The first cut is the one that surprised us. Take the same garment — an adult short-sleeve crew tee — from five different published spec sheets, and medium barely moves.

Chest flat = measured across the garment about an inch below the armhole, lying on a table. Double it for the figure around the body. Sources are each brand's own published spec sheet or size page, retrieved 22 August 2026.

One inch of spread, around the whole body, across five brands. That is tighter than most people would guess, and it points at the real mechanism. These are blanks. Printers and small labels order them by the case off a spec sheet, and a spec sheet that drifts costs the manufacturer money in returns and complaints. Consistency here is not virtue, it is commercial pressure. The number is the product.

Then it falls apart

Now change the question from "what does this shirt measure" to "who is this size for", and the agreement evaporates. Here are two live charts, both public, both retrieved the same day.

And here is the detail that does more damage than the gap itself. Neither chart tells you whether that chest column is your body or the garment. Ours is a body chart — it asks you to put a tape around yourself. ASRV's tees chart lists medium as chest 41–43 inches alongside a length of 29–32 inches. The length figure is obviously the garment; nothing on the row says which the chest figure is. If it is a body figure, their medium is cut for a notably larger chest than ours. If it is the finished garment, it is a roomy medium and the comparison changes completely.

We are not going to guess on their behalf, and that is the point worth taking away. You are routinely asked to compare two numbers that may not measure the same thing, on pages that do not tell you which. The tell is the magnitude: a figure near 20 is half a garment lying flat, a figure in the high thirties or forties is meant to go around you. If you only remember one diagnostic from this file, make it that one.

The same brand, disagreeing with itself

The last cut is the one that reframes it. Stay inside a single manufacturer and read medium off each of its published spec sheets, and the number moves further than it did across five different companies.

From 33 inches to 45½ inches around, every one of them a medium, all from one company's own spec sheets. That is a 12½ inch range. And the honest reading of it is not that the manufacturer is careless — it is the opposite. Their three crew tees are identical at 20 inches to the quarter-inch. The spread is the silhouettes doing their job. A flowy tank is supposed to skim; a drop-shoulder fleece is supposed to sit off the body. The letter is not tracking your chest. It is tracking a position within one garment's own size run, and each run starts somewhere different.

Which produces a rule worth writing down: a size letter is only comparable inside a single style. Across styles it is a sorting index, not a measurement. This is why the medium hoodie fits and the medium tank does not, from the same brand, on the same day, on the same body — and why nobody at that brand considers this a defect.

Where vanity sizing actually fits

Vanity sizing is the explanation everyone reaches for, and it is real, but it is smaller than its reputation. It means labelling a garment with a smaller size than its measurements would traditionally warrant so the number flatters the shopper. It is documented well enough to have its own economics: a 2017 paper in the Journal of Economic Behavior & Organization models why brands do it and why it is hard to stop once competitors start.

The consumer-psychology side is more interesting, because it cuts the other way. A 2014 Journal of Consumer Psychology paper examined the flip side — what happens when the size is larger than the shopper expected — and found people respond and compensate rather than shrug. The label is doing something to the person, not just to the garment. Anyone who has left a fitting room in a worse mood than they entered it already knew that.

The foundational measurement work is older and still the cleanest. In 2003, Kinley measured 1,011 pairs of women's pants across price points and label types and found inconsistency in every single size category, with more expensive pants generally running larger. That study is over twenty years old and nothing has been fixed since. What has changed is that you now buy without a fitting room, which converts a mild annoyance into a returns problem.

One honesty note on that returns problem, because it is where this topic gets sloppy. You will see confident claims that some specific percentage of apparel returns are caused by fit. Those figures circulate widely and range from roughly forty-four to seventy percent depending on the source — and the sources are overwhelmingly companies selling sizing software. We are not going to repeat a number whose provenance is a vendor's marketing page. Fit is clearly a major driver of returns. The precise share is not something we can stand behind, so we are not going to pretend otherwise.

What to do instead

The fix is not a better chart. It is changing which number you trust.

The one habit worth building is the first one. Go find the shirt you reach for constantly, lay it flat, and measure straight across from armpit seam to armpit seam. Say it comes to 21 inches. That is a 42 inch garment, and now you own a portable, brand-independent fact about what you actually like — not what your chest measures, but what a garment you enjoy wearing measures. Compare that to a flat chest figure directly. Compare it to a body chart only after you have decided how much ease that brand's silhouette implies.

It also reframes the drawer. The garments in there that do not work are not evidence about your body. They are four different design decisions that happened to share a letter.

Quick answers

Why is a medium different at every brand?

Because no rule makes the letter mean anything. The industry standards — ASTM D5585, D6240 and ISO 8559-1 — are voluntary tables of body measurements. None says what a garment labelled M must measure. Each brand picks a fit model, adds the ease its silhouette needs, and prints a letter. Two brands can attach M to chest figures five inches apart and both are within the rules, because there are no rules.

Is the chart describing my body or the shirt?

Often the page does not say, which is the single biggest source of confusion. Use magnitude as the tell. A figure near 20 inches is the garment lying flat — half the chest, so double it. A figure in the high thirties or forties is meant to be you. Our chart is a body chart. Most streetwear blanks publish garment charts. Plenty of brands publish something in between and label neither column.

Are blank tee brands really more consistent?

Measurably, yes. Five published spec sheets for adult short-sleeve crew tees put medium between 20 and 20.5 inches flat — one inch of spread around the body. Blanks are ordered by the case off a spec sheet, so drift is a commercial problem for the manufacturer. Fashion brands compete on how a garment looks, so drift is a design decision rather than an error.

Does one brand at least stay consistent with itself?

Within a style, generally yes. Across silhouettes, no. One manufacturer's own spec sheets put medium at 16.5 inches flat on a flowy tank and 22.75 on a drop-shoulder fleece — 12.5 inches of difference around the body. Its three crew tees were identical at 20 inches. The letter tracks a position in that style's size run, not your chest.

What is the fastest thing I can do before I buy?

Measure a garment you already own and love — flat, armpit to armpit — and keep that number in your phone. It beats your body measurement because it already contains the ease you like. Then match the basis before you compare. If a brand publishes no numbers at all, treat that as information about the brand.

Keep reading

• which measurement do i actually go by. — File 11: how to read a single chart when your chest and waist disagree
• what's my size, then. — File 02: why the three-size body is the majority case
• why don't men's shirts ever fit me right. — File 10: the pattern block underneath all of this
• what does gsm actually mean. — File 09: the other number on the page that nobody explains
• why won't they just make more. — File 08: why the size you wanted is gone

---

References

A note on our own numbers. The comparisons here are arithmetic on published charts, not a peer-reviewed study: we doubled flat-chest figures where the spec sheet states the measurement is taken laid flat, and left body figures as printed. Charts change without notice, so every figure is dated 22 August 2026 and linked so you can check it. We dropped one spec sheet whose column mapping was ambiguous rather than guess at it. Where two brands' charts may not be measuring the same quantity we have said so instead of resolving it in our favour, and the one comparison that reflects badly on us — our medium being the smallest in the set — is in the chart rather than the footnotes. Product education from the development room, not fit advice for any individual body.

feel scared. feel so alive.