Garment measurements versus body size charts, and how to tell which you are reading

Two size charts can carry the same column headings and mean opposite things. One lists the body the garment is meant to fit. The other lists the garment itself, lying flat on a table. The difference between them is the ease, which on a woven top is 5cm to 10cm and on something cut loose can be 20cm or more. Read a garment chart as a body chart and you will buy two sizes too big. Read a body chart as a garment chart and you will buy something you cannot do up. Here is how to tell them apart in about ten seconds.

The three tells that identify a body chart

Body charts give a range per size. A size M reading bust 88cm to 92cm is describing people, because a garment has one chest measurement and not a range of them.

Body charts list only body dimensions: bust, waist, hip, sometimes height and inside leg. If a chart has a column for length, shoulder, sleeve or hem width, it is describing a garment, because nobody's body has a hem.

Body charts are also usually generic across a whole shop, identical on every product page. A chart that changes from item to item is a garment chart, and a chart that is the same on a coat and a bralette is a body chart.

Garment charts, by contrast, give a single number per size, often alongside a tolerance note, and the numbers are larger than any body that size would be. A size M listed at chest 104cm is not a person's chest. It is a garment with 14cm of ease in it.

Half measurements, which cause the worst errors

Many garment charts give half measurements, taken across the flat garment rather than around it. Pit to pit, half chest, half waist and half hem all mean one side only, and you double them to get a circumference.

The giveaway is magnitude. If a size L shows chest 52cm, that is half of 104cm and not the measurement of anything else. If it shows 104cm, it is already a circumference. Numbers in the 40s and 50s for a chest are halves. Numbers in the 90s and above are full.

Get this wrong in either direction and you are out by a factor of two, which is the single largest error available in the whole process. When a chart gives no units and no explanation, that magnitude check is the only thing you have, and it is reliable.

Length measurements are never halved. Length, shoulder, sleeve and rise are all single straight runs, so take them as given.

How much ease to add, by fabric

A woven garment with no stretch needs about 5cm over your body measurement at the governing point for a close fit you can sit down in, 8cm to 10cm for a normal fit, and 15cm or more to read as loose. Less than 5cm and a woven bodice pulls at the closure.

A knit with elastane can sit at your measurement, and a design meant to be held under tension can sit several centimetres under it. This is the only case where a garment measurement smaller than your body measurement is correct rather than a mistake.

Velvet behaves like a woven with a little extra bulk. The pile occupies real space, so a velvet piece at a given flat measurement feels slightly closer than a flat woven at the same number. Add a centimetre or two to what you would otherwise want.

Lining takes up room too, and so does anything with a facing or interfacing at the closure. A lined garment at the same flat chest measurement as an unlined one is the tighter of the two.

Tolerance, and why a 1cm difference is not a decision

Garment manufacturing tolerance is normally 1cm to 3cm on a circumference, more on a longer dimension. Two items of the same size from the same factory can differ by that much and both be correct.

So a chart is a target rather than a promise. If two sizes are within about 2cm of your ideal at the governing measurement, the chart is not telling you which to pick and you should decide on something else: fabric, construction, which direction of error you can live with.

This is the reason you should not add a precise fudge factor to a chart. People calculate to the half centimetre from a figure that was never that accurate, and the arithmetic feels like control. The useful precision is in your own body measurements, where you can be accurate, not in the garment's published numbers.

Measuring a garment you already own, which beats every chart

Find the item in your wardrobe closest in cut and fabric to the one you are buying, one that fits the way you want this one to fit, and measure it flat. Then compare numbers to numbers. This removes ease, grading and chart interpretation from the problem entirely.

Lay it on a hard flat surface, buttoned or zipped, smoothed but not stretched. Chest: straight across from one armpit seam to the other, 2.5cm below the armhole, and double it. Length: from the highest point of the shoulder seam at the neck, straight down to the hem. Shoulder: across the back from one shoulder seam point to the other. Sleeve: from the shoulder seam to the cuff edge. Waist and hem: straight across at the narrowest and widest points, doubled.

Write those down beside your body measurements and keep both lists. They are the most useful piece of paper in buying clothing online, and they take ten minutes once.

This is also the only sane way to judge something cut deliberately large. A coat cut big on purpose is not trying to match your chest, so comparing its chest number to yours tells you nothing useful. Its shoulder and sleeve are the numbers that decide whether it works.

What we publish, and what to do when a page is short

Our clothing pages state the sizes offered before you buy. Where we have the supplier's flat garment measurements we give them, and where we do not, we leave the chart off rather than copying across a generic one that would be wrong for that item.

Our printed cotton, including the printed tees, runs in standard United States sizes from S to 3XL and behaves as you would expect, which makes it a useful calibration point. Pieces that come from overseas suppliers do not run to the same block, and we have written separately about why the same letter means different measurements in different places.

If a page does not give you what you need, email us with the item and your governing measurement and we will read the supplier's own figures for that exact piece. We do not take returns for a size chosen wrongly, so that email is the only step in the process that genuinely reduces your risk. A piece that is close through the hip is exactly the kind of thing worth asking about first.

Questions people actually ask

How do I know if a size chart is body or garment measurements?

A body chart gives a range per size and lists only bust, waist and hip. A garment chart gives one number per size, includes length, shoulder or sleeve columns, and its figures are larger than any body that size would be. A chart that is identical on every product page is a body chart.

What does pit to pit mean on a size chart?

The distance straight across the flat garment from one armpit seam to the other, which is half the chest circumference. Double it to compare with your own bust measurement. A chest figure in the 40s or 50s of centimetres is a half measurement; one in the 90s or above is already a circumference.

How much ease should a top have over my bust measurement?

About 5cm for a close fit you can still sit in, 8cm to 10cm for a normal fit, and 15cm or more to read as loose. A knit with elastane can sit at your measurement or slightly under. Velvet and anything lined take up a little extra room.

Why is my garment a different size from the chart?

Manufacturing tolerance on a circumference is normally 1cm to 3cm, and more on a length. Two garments of the same size can differ by that much and both be within spec. Treat a chart as a target, and do not choose between sizes on a 1cm difference.