Why are clothing sizes so inconsistent between brands?

Because there is no mandatory standard requiring them to agree. The US had a voluntary government standard for women's sizing, Commercial Standard CS 215-58, and it was withdrawn in 1983. What replaced it is optional ASTM guidance that individual brands can adopt or ignore, so a size 8 is whatever the brand's own pattern maker decided it was.

Was there ever an official US clothing size standard?

Yes. The National Bureau of Standards published Commercial Standard CS 215-58 in 1958, built from a 1939 to 1940 USDA survey that measured nearly 15,000 women. It was updated in 1970, then downgraded to voluntary status and withdrawn in 1983.

What replaced the old US sizing standard?

A set of voluntary ASTM International standards, including D6829 for misses' body measurements and D6960 for women's plus sizing. Brands can reference them in their pattern making, but nothing requires it, and nothing checks whether a labeled size actually matches the chart.

How do I know what size will actually fit me?

Ignore the letter or number on the tag and compare flat garment measurements, chest width, body length, sleeve length, against a piece you already own and like the fit of. Most brands publish these measurements on a size guide page; if they don't, ask before you buy.

Why clothing sizes make no sense

In the summer of 1939, the federal government put nearly five hundred people to work with tape measures. The project ran under the Works Progress Administration, the Depression-era jobs program better known for building bridges and painting murals, and its assignment was narrower and stranger: measure American women, one at a time, until someone finally knew what shape they were. Between July 14, 1939, and June 30, 1940, relief workers took 58 measurements on each of 14,698 women, from bust and waist down to wrist and ankle. Ruth O’Brien and William C. Shelton of the USDA’s Bureau of Human Nutrition and Home Economics published the results in 1941 as Women’s Measurements for Garment and Pattern Construction. It remains one of the largest anthropometric surveys of women’s bodies ever conducted in the United States.

The point of the survey was not curiosity. Pattern makers and garment manufacturers had been sizing clothes by guesswork and inherited convention for decades, and it showed: a size 16 from one maker had no relationship to a size 16 from another. O’Brien and Shelton’s data was meant to fix that by finding, in their own words, “a few important measurements from which all the others can best be predicted,” so the trade could build sizes around actual bodies instead of a tailor’s memory. The idea was that a size, once defined against real measurements, would mean the same thing wherever it was printed.

The standard that existed, and how it died

That data eventually became something with legal-sounding teeth: Commercial Standard CS 215-58, published by the National Bureau of Standards in 1958 and titled Body Measurements for the Sizing of Women’s Patterns and Apparel. It assigned numeric sizes to specific bust, waist, and hip measurements, organized around body types rather than one universal scale. It was updated once, in 1970. For a couple of decades, the US had the closest thing it has ever had to a government-defined women’s clothing size.

A commercial standard, though, was a voluntary reference document that manufacturers could choose to follow, the way a building code recommendation differs from a building code requirement. Compliance drifted as manufacturers found that a size 8 that measured smaller than the chart said sold better than a size 8 that measured true. The standard was downgraded further, and by 1983 it had been withdrawn altogether. No replacement with government backing has existed since, and no comparably sized national body-measurement survey has been run to replace the 1939 to 1940 data that started the whole project.

What sits in that vacuum now is a set of voluntary standards from ASTM International, the private standards organization that also writes specs for concrete and bicycle helmets. ASTM D6829 covers misses’ body measurements; ASTM D6960 covers women’s plus sizing. Any brand can reference these tables when it builds a pattern. Nothing requires it to, and nothing checks the label against the tape measure afterward. A brand is free to call a garment a medium regardless of what its own spec sheet says a medium measures, and plenty do, because a size label is marketing copy wearing a number.

Vanity sizing is the predictable result of no enforcement

Once compliance became optional, the incentive ran one direction. A shopper who fits into a smaller labeled size on a store’s website is more likely to buy and less likely to return the item, so a brand chasing conversion has a reason to make its size 8 run larger than the last decade’s size 8 while leaving the tag alone. Nothing in the current system stops that drift, because nothing with authority is checking a label against a tape measure anymore. The result is a size label that measures a brand’s marketing position more reliably than it measures a body.

This is not the same failure as inconsistent quality control. A badly cut garment is a mistake. A size chart that expands over the years while keeping the same tag numbers is a decision, made by someone in a design meeting who was looking at a return rate, not a body. Both problems land on the customer the same way: the tag stops being useful information.

One corner of the industry never gave up inches

Menswear tailoring solved this problem by refusing to have the argument in the first place. A suit jacket is sold as a 40 or a 42, and that number is a chest measurement in inches, not an arbitrary rank. A 42 long tells a tailor the jacket’s chest and its sleeve length before it leaves the rack. The system survives because it was never translated into a letter that a brand could redefine at will. Nobody has to trust a retailer’s judgment about what “medium” means this season, since the number on the tag is the same unit a tape measure produces.

Letter and number sizing for shirts, dresses, and pants gave that up somewhere along the way, trading a measurement for a rank: small, medium, large, or a size 8 that means whatever the pattern maker’s block says it means. A rank is faster to print on a tag and easier to make aspirational, small sounding better than it should, but it carries no information a shopper can act on across brands. Suiting kept the inches because tailors, unlike marketing departments, need the number to be true.

What the tag can’t tell you and the tape measure can

None of this means sizing is arbitrary within one company’s own catalog. A brand’s fit block, the internal set of measurements it builds every garment around, is usually consistent with itself even when it disagrees with everyone else’s. The instability is between brands, not within one. Which is why the only number worth trusting across a purchase is the one nobody can redefine without you noticing: the garment’s own flat measurement, chest width pit to pit, body length shoulder to hem, sleeve length, taken with a tape measure against something already hanging in the closet.

That comparison takes two minutes and no math. Lay a shirt you already like flat on a bed, measure armpit to armpit for the chest and high shoulder point to hem for the length, and hold those two numbers against any size chart before you order. A chart that lists only “medium” without inches attached is giving you a guess with a brand name on it.

DRESS publishes exactly that on its own size guide: flat chest, length, and sleeve measurements in inches for every garment, next to the letter size rather than instead of it. The letter is a shortcut. The inches are the fact.

Buy by the tape measure, and a size label goes back to being a convenience instead of a guess. Ruth O’Brien’s crew spent a year with 14,698 women and a measuring tape trying to make that convenience trustworthy again. Eighty-five years on, the tape measure is still the part that works.