Standard clothing sizes trace back to the American Civil War, when the Union Army needed to outfit hundreds of thousands of soldiers without custom tailoring each uniform. That wartime experiment produced the first large-scale body measurement data and launched a sizing system that evolved over the next 160 years into the labels you see on store racks today. The path from battlefield to fitting room involved military logistics, government-funded science, industrial standards bodies, and, more recently, 3D body scanning technology.
The Civil War Created the First Size Charts
Before the 1860s, nearly all clothing was either homemade or produced by local tailors who measured each customer individually. When the Civil War erupted in 1861, both the Union and Confederate armies faced an unprecedented problem: how do you make clothes that fit thousands of bodies without tailoring each one?
The Union Army’s Quartermaster Department solved this by collecting measurements from thousands of soldiers, recording chest, waist, and inseam dimensions, then producing uniforms in a range of sizes based on averages. This was one of the first large-scale efforts to categorize human body types for clothing production. The uniforms weren’t perfect fits, but they were good enough to clothe an army, and the approach was revolutionary.
After the war ended, clothing manufacturers applied these same sizing principles to civilian garments. The military data gave rise to the modern concept of small, medium, and large, along with numerical sizing for suits, shirts, and trousers. There was a significant limitation, though: the entire system was built on male soldiers’ bodies. Women’s clothing remained largely custom-made or loosely fitted for decades afterward.
A 1939 Government Study Tackled Women’s Sizes
By the early twentieth century, ready-to-wear clothing for women was growing fast, but sizing was chaotic. Manufacturers each used their own systems, and a dress labeled one size at one store could fit completely differently at another. The U.S. Department of Agriculture’s Bureau of Home Economics set out to fix this by conducting a massive anthropometric study of American women.
The research project, led by Ruth O’Brien and published in 1941, measured a large group of women and performed detailed statistical analysis on the results. One of its most important findings was counterintuitive: girth measurements have little relationship to height. Women with a 46-inch bust were no taller, on average, than women with a 32-inch bust. Women standing 5 feet 9 inches had no larger busts, on average, than women who were 5 feet tall. This meant that bust measurement alone, which was commonly used to size women’s dresses, couldn’t predict how long a garment needed to be.
The researchers concluded that a combination of one vertical measurement and one horizontal measurement was necessary to accurately predict a woman’s full set of body dimensions. Height worked best as the vertical measurement, and weight turned out to be a better predictor of all girth measurements than any single girth measurement like bust or waist. A height-weight combination, they argued, would be the ideal basis for classifying women’s body types and establishing a standard sizing system.
The First National Standard and Its Limits
In 1958, the U.S. government published Commercial Standard CS 215-58, the country’s first voluntary standard for women’s clothing sizes. It drew heavily on the O’Brien study data, assigning specific bust, waist, and hip measurements to numbered sizes. For the first time, a “Size 12” or “Size 16” was supposed to mean the same thing regardless of which manufacturer made the garment.
The standard had problems from the start. The women measured in the original study were predominantly young and white, which meant the data didn’t reflect the full diversity of American body shapes. The system also assumed proportional relationships between measurements that didn’t hold up across the real population. Manufacturers found the standard too rigid, and compliance was always voluntary, so many companies quietly ignored it or modified it to suit their own customers. The government eventually withdrew the standard in 1983, leaving the industry without an official sizing framework.
How Vanity Sizing Changed the Numbers
Once there was no binding standard, brands were free to adjust their size labels however they wished. Most adjusted them downward. A garment with a 28-inch waist might have been labeled a Size 10 in 1960 but gradually became a Size 6 or even a Size 4 as brands discovered that customers felt better buying a smaller number. In the 50 years between 1958 and 2008, a U.S. Size 8 expanded by up to 6 inches in actual measurements.
This drift happened unevenly across brands, which is why you can wear three different sizes depending on where you shop. Budget retailers and luxury brands tend to run at opposite ends of the vanity spectrum, with high-end labels often using the most generous labeling. The result is a system where the number on the tag tells you very little about the actual dimensions of the garment.
Modern Standards and 3D Body Scanning
Today, the closest thing to an official sizing framework in the U.S. comes from ASTM International, which publishes voluntary standard tables of body measurements. Their standard for adult women, for example, covers misses figure types in sizes 00 through 20, listing the full range of body measurements manufacturers should use when developing patterns. These tables are updated to reflect current population data, but adoption remains voluntary.
The data feeding these standards has gotten dramatically more precise. The CAESAR project (Civilian American and European Surface Anthropometry Resource), a multinational survey conducted in the early 2000s, used 3D body scanners to capture detailed surface measurements of thousands of civilians across North America and Europe. The scanners proved highly reproducible, with most measurements accurate to within 5 millimeters, making them as good as or better than traditional manual measurement techniques. This kind of data allows for far more nuanced sizing categories than anything available during the Civil War or even the mid-twentieth century.
Three-dimensional scanning is now used not just for population surveys but increasingly in retail. Some companies use scan data to develop better grading systems (the formulas that scale a pattern up or down across sizes), while others offer personalized fit recommendations based on individual body scans. The technology exists to move beyond standard sizes entirely, but the economics of mass production still favor a system of discrete size categories, however imperfect they remain.
Why Sizes Still Don’t Work Well
The fundamental challenge identified in 1939 has never been fully solved: human bodies vary in too many independent dimensions to be captured by a single number. Height doesn’t predict girth. Bust size doesn’t predict waist size. Two people who wear the same pants size can have completely different torso proportions. Any system that reduces this complexity to “Size 8” or “Medium” is going to be a rough approximation.
Add in vanity sizing, brand-specific fit preferences, and the fact that no global standard exists (a European 38 is not the same as a U.S. 8, even though conversion charts suggest they are), and the modern sizing landscape is arguably more confusing than it was a century ago. The measurement tools have gotten better. The willingness to use them consistently across an industry has not.