How to find more shirts like the one you already own

Measure the shirt you already own while it lies flat, write down eight numbers, then compare those numbers against published garment measurements. Shirts within about half an inch on chest and length will wear much the same way.

Work from the shirt, not from its label. Lay the shirt you already like flat, take eight measurements, then look for garments whose published measurements sit within about half an inch of yours. Matching numbers to numbers is the whole method, and it is the thing the shirt scanner automates.

What is the short version of the method?

Choose the reference shirt — the one that already hangs the way you want, not the newest one in the drawer.
Lay it flat and square it up — side seams straight, collar fastened if it has buttons, no part of the fabric pulled tight.
Take eight measurements — chest, body length, shoulder, sleeve length, sleeve opening, bicep, neck opening and hem.
Record them to the nearest quarter inch — anything finer disappears inside manufacturing tolerance.
Compare against published garment measurements — the brand's own spec sheet, the blank manufacturer's size chart, or a catalogue that already holds those numbers.
Read the gap in inches — a shirt 0.5" wider in the chest reads as slightly roomier, and a shirt 2.0" wider is a different garment.

Which eight measurements should I record?

Eight points describe a shirt closely enough to shop with. Chest and body length do most of the work. The other six explain why two shirts with the same chest still feel different once they are on.

Point	How to take it	What it decides
Chest, pit to pit	Straight across the garment, one inch below the armhole seam	Looseness through the body; the figure most brands publish
Body length	Highest point of the shoulder straight down to the hem	Whether the shirt tucks, sits at the belt, or covers the hip
Shoulder	Seam to seam across the back	Where the sleeve head lands on your arm
Sleeve length	Shoulder seam out to the sleeve hem	Whether a short sleeve stops above or below the elbow
Sleeve opening	Across the sleeve hem, laid flat	Most of the visual difference between a boxy tee and a fitted one
Bicep	Across the sleeve, one inch below the armhole	Where a shirt that fits the chest can still bind
Neck opening	Across the inside of the collar, laid flat	How the shirt sits at the throat, independent of every other figure
Hem width	Across the bottom opening, laid flat	Whether the body tapers or hangs straight

If you have not taken these before, how to measure a shirt runs through the sequence, and pit to pit covers the single most useful one. Sleeve length has its own page because brands disagree about where that measurement starts.

How close does a match have to be?

Read the difference in inches and sort it into three bands. All three assume you are comparing flat garment measurements with flat garment measurements, never a garment figure against a body figure.

Chest difference	How to treat it
0" to 0.5"	Inside the tolerance band of the specs themselves. The two garments may not differ at all in practice.
0.5" to 1.0"	A real difference, but a wearable one. Direction matters: wider reads looser, narrower reads closer to the body.
Over 1.0"	A different fit. Past 2.0" you are comparing a fitted cut with a relaxed one, whatever the labels say.

What do published chest numbers look like across common blanks?

These are the flat chest figures ShirtMath holds for eight widely stocked blank styles. They show why the method works: the numbers are public, they are specific, and they disagree with each other in ways the size letter hides.

Style	S	M	L	XL
Gildan G500 (unisex)	18.0"	20.0"	22.0"	24.0"
Bella+Canvas 3001C	18.0"	20.0"	22.0"	24.0"
Hanes 5250T	18.0"	20.0"	22.0"	24.0"
Champion T525C	18.0"	20.0"	22.0"	24.0"
Gildan 5400 (long sleeve)	18.0"	20.0"	22.0"	24.0"
Next Level 3600 (fitted)	19.0"	20.5"	22.0"	24.0"
Comfort Colors C4017	18.5"	20.5"	22.5"	24.5"
Gildan G500L (ladies cut)	16.5"	17.5"	19.0"	21.0"

Read the Medium column. Five unisex styles publish 20.0", and the fitted Next Level 3600 and Comfort Colors C4017 both publish 20.5". Body length at Medium is 29.0" for every unisex style here except the C4017, which is the longest at 29.5". So two Mediums can agree exactly on chest and still differ in length, which is exactly why one measurement is not enough. More chest figures sit in the chest measurement data table, and which blank runs biggest ranks the styles against each other.

One row in this table is widely misquoted. The 22.0" chest people attach to Comfort Colors C4017 at Medium is a size L figure on the unisex ladder, and the C4017's own Large is 22.5". Read the published sheet and Medium is 20.5" against the 20.0" of Gildan G500: half an inch, not two, and well inside the ±0.8" tolerance a garment-dyed style carries. The length difference at that letter is half an inch as well. If a shirt you own feels roomier than a Gildan of the same letter, check the actual numbers before blaming the brand.

Why does the size letter not travel between brands?

Because the letter is a label, not a measurement. The clearest example in this data set is not a brand difference at all, it is a cut difference inside one brand. A ladies Gildan G500L Medium publishes 17.5" across the chest. The unisex Mediums publish 20.0" to 20.5". That is a gap of two and a half to three inches on the same letter, and it reaches 3.5" at Large. Someone who owns a G500L Medium and buys a unisex Medium is buying a different garment.

This is the mechanism behind vanity sizing, and it is also why the same size in a different brand is an unanswerable question until you put numbers on both garments. Note also that some brands publish body measurements, meaning the measurements of the person, rather than the flat measurements of the garment. Mixing the two produces nonsense; body measurement versus garment measurement sets out the difference.

Where does matching by measurement break down?

Four limits, stated plainly.

Catalogue coverage — a comparison can only return garments whose measurements are actually held. If a brand does not publish flat measurements and is not in the catalogue, no amount of accuracy on your side will surface it. This is the real ceiling on the method today.
Fabric and knit — two shirts can share a chest figure and hang differently because of weight, knit and finish. Published specs describe dimensions, not drape.
Tolerance — roughly ±0.3" to ±0.8" on published figures, wider on garment-dyed styles such as C4017.
The shirt you own has a history — washing and wear change a garment. You are measuring what you own now, which is the right thing to measure if you like how it fits now, and the wrong thing if you are trying to reproduce how it fitted when new. In that case use the published spec for the style instead.

What can ShirtMath do today?

The live capability is narrow and worth stating exactly. Photograph a shirt laid flat next to a scale card, and the scanner returns flat measurements for that garment and compares them against published garment measurements across the catalogue. Roughly half of scanner readings land within an inch of a tape measure. Sleeve opening and bicep are the least reliable points. Anything the scanner cannot read confidently is left blank rather than guessed, so blank fields in a result are a design decision, not a failure to display.

A tape measure is still more accurate than a photograph. If you own the shirt and own a tape, measure it by hand and type the numbers in. The scanner exists for speed, not for precision beyond a tape. Shoe sizing, eyewear measurement and body scanning are not part of what runs today. The methodology page sets out how readings are produced and checked, and saved garments live in your closet so that a future comparison starts from numbers you have already taken. Results appear under your matches.

Common questions

Do I need the original shirt's brand and style number?

No. The measurements are the identity that matters, and a shirt with no label at all can still be matched. A surviving label helps, though: blank style numbers such as G500 or 3001C map straight onto a published specification, which saves you measuring at all.

What if the shirt I own has stretched or shrunk?

Measure it as it is now. The garment in your hand is the one you actually wear, so its current numbers are the ones to shop against. If you want the shape it had when new, use the published spec for that style as your reference instead of the worn garment.

Is chest width alone enough to find a similar shirt?

It gets you most of the way for a T-shirt, and no further. At Medium, body length is 29.0" across the unisex styles here and 29.5" on Comfort Colors C4017, so shirts that agree exactly on chest can still differ by half an inch in length.

Does a half inch really matter?

Usually not on its own, because half an inch sits inside the manufacturing tolerance of the published spec. It starts to matter when several half inches point the same way, for example a shirt that is half an inch wider in the chest and half an inch shorter in the body.

Can I do this without a scale card?

Yes. A tape measure and a flat table are enough, and they are more accurate than a photograph. The scale card is only needed when you want the measurements read from an image rather than taken by hand.

Should I match a woven shirt to a knit T-shirt?

Not on numbers alone. An identical flat chest behaves differently in a stable woven and in a stretchy knit, so compare like with like where you can, and treat a cross-fabric match as a starting point rather than a result.

What if no shirt in the catalogue matches my numbers?

Then coverage is the limit, not your measurements. The eight figures you recorded are portable: take them to any brand that publishes flat garment measurements and do the comparison yourself. Nothing about the method depends on a particular catalogue.

How accurate is measuring by photograph?

Roughly half of scanner readings land within an inch of a tape measure, and sleeve opening and bicep are the weakest points. Readings the scanner is not confident about are left blank. Use a photograph for speed and a tape when the number has to be right.

Where to go next

This page sits in Answers, the plain-question section of the site. These are the pages that follow on most directly.

How to measure a shirtThe full eight-point sequence, with where each measurement starts and stops. Vanity sizingWhy the same letter names different garments across brands and cuts. Chest measurement dataPublished flat chest figures across blank styles and sizes. All answersShort, direct answers to the sizing questions people actually ask.

How to find more shirts like one you own