Ask five people what size they wear and you will often get five answers that have nothing to do with their bodies and everything to do with which brands happen to fit them. That is not a coincidence. For most of the twentieth century, American women's clothing was cut against a shared federal reference. That reference is gone, and it has been gone for over four decades, which is the real reason a medium in one store can be a large in another with no consistent rule to predict which.
The size printed on a tag used to point to something. Now it mostly points to a marketing decision made inside one company. Once you know that history, chasing a number stops making sense, and a tape measure becomes the only tool that tells you anything reliable.
There used to be an actual standard
In 1939, the United States Department of Agriculture ran a nationwide study called Women's Measurements for Garment and Pattern Construction, recording weight and fifty eight body measurements from nearly fifteen thousand women across seven states. The results were shared with the clothing industry for comment, and in 1958 the findings became a federal reference called Commercial Standard CS215-58.
That standard was not perfect even at the time. It built its size range from eight to thirty eight around a single hourglass proportion, using bust measurement as the anchor and adding a short, regular, or tall notation along with a plus or minus mark for hip and waist variation. Bodies that did not match that one proportion were already poorly served. But at least every company selling a size ten was working from the same rough target, so a size ten from one label had a reasonable chance of resembling a size ten from another.
1983 is the year sizing stopped meaning anything shared
CS215-58 was revised and made fully voluntary in 1970, and by 1983 the Department of Commerce withdrew the standard altogether. No mandatory replacement ever arrived. From that point forward, each retailer was free to define its own numbers, and many used that freedom to make their sizes run larger than the old federal scale while keeping the same familiar labels, a shift most sizing historians point to as the true starting line for what people now call vanity sizing.
The incentive behind that drift is not mysterious. A shopper who tries on a smaller number and has it fit comfortably tends to feel good about the purchase and the brand, regardless of what the tag says. Multiply that decision across an entire industry with no shared reference pulling everyone back toward a common scale, and you get exactly what exists now: labeled sizes that have wandered so far apart between companies that the same person can measure identically and still need three or four different numbers depending on where they shop.
Why your past purchases stop being useful information
Once you understand that there is no shared reference behind the label, it follows that remembering your size from one brand tells you very little about a different brand. This is the single biggest mistake in fast online shopping: ordering a familiar number from an unfamiliar label and being surprised when it does not match.
- Grading differs by brand. Two companies can both offer a size eight and cut it around entirely different bust, waist, and hip measurements.
- Category changes the target body. A single retailer may size denim against a different reference than the one it uses for blazers, so even within one label the number is not fixed.
- Country of origin changes the scale. A size that reads correctly on a European site can shift several numbers once it lands on an American size chart.
- Fabric stretch changes what a printed size tolerates. A rigid woven garment and a garment with added elastane can both claim the same size while fitting completely differently.
It is not only a women's clothing problem anymore
Men's sizing has its own version of this drift, since men's tags are often built around neck, chest, or waist inches rather than an arbitrary number, which sounds more objective but still varies by how generously each brand cuts the garment around those measurements. Children's sizing adds another layer, since age-based labels have to guess at a growth curve that any individual child may run ahead of or behind. None of these systems failed for the same reason CS215-58 failed, but all of them share the same lesson: a label is a starting guess from one company, not a universal fact about a body.
It helps to remember that the 1958 standard was already a compromise built around one body shape, so its disappearance did not erase a perfect system, it erased a shared reference point that at least gave shoppers a common baseline to argue with. What replaced it was not a better system, it was dozens of separate systems with no obligation to agree with each other or with the one that came before.
Measure once, then compare that number to every chart
The fix is not complicated, it is just a habit most people never build. Take your own bust, waist, and hip measurements with a soft tape, standing naturally, and write them down somewhere you can find again. That short list becomes more useful than any size you have ever worn, because it does not change depending on who cut the garment.
- Measure in inches or centimeters, not in a size you assume you already know. Wrap the tape around the fullest part of the bust, the natural waist, and the widest part of the hips.
- Open the size chart before you order, every time. Most retailers publish one on the product page, even if it is buried near the bottom.
- Match your actual measurements to the chart's ranges. Ignore the label letter or number until you have already found which row fits your body.
- Note the fabric content next to your decision. A garment with stretch can allow you to size down from the chart; a rigid weave usually cannot.
- Keep a short brand log. Once you learn that one retailer runs small and another runs generous, that pairing becomes durable knowledge you can reuse.
None of this requires giving up on brands you already like. It just moves the decision away from a number that stopped being reliable in 1983 and back toward the two measurements that have never lied to anyone: the tape, and the chart in front of you.
Start with the one garment you return the most
Pick the clothing category that has disappointed you most often, whether that is trousers, blazers, or fitted knitwear, and measure yourself properly before your next order in that category. Compare that number against the specific brand's chart rather than the size you wore last time somewhere else. A five-minute habit now saves a return label later, and it works regardless of how any single company chooses to label its garments.