To grade shrinkage on used cotton sweaters, lay the sweater flat, measure chest, body length, and sleeve length, and compare each to a reference: the tag's size spec or a known unshrunk sibling. Under about 2% loss is normal wear. Past about 4% in length or sleeve, or any twisting, it is a defect you disclose and grade.
Here is the return we keep seeing. A seller lists a size L cotton crewneck as Excellent. The buyer gets it, holds it up, and it fits like a M. The seller never measured it. The buyer opens a not-as-described case, and the seller eats return shipping on a sweater that sold for $28. A tape measure and one sentence in the listing would have prevented it.
The percentages in this article are working tolerances we use as examples, not industry standards. Adjust them to your category, but pick numbers and apply them the same way every time.
Why Shrinkage Is a Measurable Defect, Not an Opinion
Cotton knit shrinks in two directions, and not equally. Length usually loses more than width, because the loops relax and tighten along the wale. Sleeves lose length fast. A sweater can still read as the right size across the chest while the body is two inches short. Buyers feel that difference immediately.
Shrinkage touches several of the five grading factors:
- Fabric Condition: a hard, dense, felted hand means the fibers have tightened past normal wash wear.
- Structural Integrity: twisted side seams, a ribbed hem that no longer sits flat, and a collar that cups against the neck.
- Cosmetic Appearance: distorted stripes or pattern repeats, and a boxy, compressed silhouette.
- Functional Elements: a neckline too tight to pull over a head, or cuffs that ride up the forearm.
This is why we treat it like any other item in the defect taxonomy. It has a test, a threshold, and a disclosure. If you want the same method applied to wool, we covered it in our post on how to tell if a wool sweater has shrunk. Cotton behaves differently, so the tolerances below are set for it.
Pick Your Reference Before You Measure
A measurement means nothing without something to compare it to. You have three options, ranked from strongest to weakest:
- A sibling garment. Same brand, same style, same size, never washed or hardly worn. Best reference, rarely available.
- The manufacturer's size chart. Many brands publish flat measurements per size. Check the brand's site or an archived listing.
- Tag-size comps. Measure two or three other used sweaters of the same brand and tag size and take the typical range. It is weaker, because used comps may have shrunk too, but the pattern shows up fast.
If you have no reference at all, say so. Report the measurements and let the buyer judge. That is honest, and it still prevents the dispute.
The Flat Measurement Protocol
Do this on a clean, flat surface with the sweater smoothed but not stretched. Pulling it by even half an inch ruins the comparison. Here is the sequence:
- Let it rest. Lay the sweater flat for a few minutes so the knit settles into its natural shape.
- Measure chest. Armpit to armpit, straight across, one side of the garment. Double it only if your reference is a circumference.
- Measure body length. From the high point of the shoulder, beside the collar, straight down to the bottom of the hem.
- Measure sleeve length. From the shoulder seam (or the collar center on raglans, matching your reference method) to the end of the cuff.
- Check the hem and cuffs. Do the ribbed bands lie flat or flare outward? Flaring hems often signal a body that has pulled up.
- Run the twist test. Lay the sweater flat, line up the side seams, and see whether they spiral toward the front. Spiral means the knit has relaxed unevenly.
- Compute the loss. For each dimension: (reference minus measured) divided by reference, times 100. Record all three.
- Photograph the tape. One shot with the tape across the chest and one down the length, so the numbers are on record.
The whole routine takes about three minutes per sweater once you have done it a dozen times. Write the numbers on a card or in your notes before you move on to the next item.
Set a Tolerance That Justifies a Tier
Every cotton sweater loses a little. A tolerance tells you where normal ends and a defect begins. The table below shows example thresholds. They are a starting point for your own consistent standard, not a universal rule.
| Measured loss (worst dimension) | Other signs | Shrinkage call | Tier effect |
|---|---|---|---|
| Under 2% | Flat seams, even hem | Normal wash wear | No effect; grade on other factors |
| 2% to 4% | Slight hem flare, no twist | Minor shrinkage | Cap at Very Good if it affects fit; disclose |
| 4% to 7% | Short sleeves or body, possible mild twist | Clear shrinkage | Good; disclose with measurements |
| Over 7% | Twisted seams, dense hand, tight neck | Significant shrinkage | Fair; lead the listing with it |
Two rules keep the table honest. First, grade on the worst dimension, not the average. A buyer who can't wear the sleeves does not care that the chest is fine. Second, a sweater that sits in a high tier on every other factor still gets capped. A pristine, unpilled, odor-free cotton crewneck that now measures a size smaller is not Excellent. It is a good sweater in a different size, and it should be listed that way.
Note the interplay with the other factors. If shrinkage also left the fabric felted and stiff, Fabric Condition drops on its own. If the sweater has a stain and a shrunk body, grade each factor and let the lowest weighted result drive the tier. Don't let one clean factor pull a damaged garment upward.
When Shrinkage Is Intentional or Doesn't Count
Not everything that measures short is a defect. A few cases to separate:
- Cropped or boxy cuts. If the brand cut it short, your reference already says so. Measure against the right chart.
- Pre-shrunk or garment-washed cotton. These start soft and slightly compressed. The size chart usually reflects that.
- Vintage sizing. A 1990s Large often measures like a modern Medium. That is a sizing drift, not shrinkage. Do not blame the wash.
- Stretch, the opposite problem. A sagging, elongated body with a loose neck is a different defect. Measure it the same way and report negative loss.
If you can't tell which one you have, the measurements still do the work. List them, state the reference you used, and let the buyer compare to their own garments.
Disclosure Wording for Shrunk Knits
Disclose whenever the loss is over your minor threshold, or whenever the tag size will mislead a buyer. Put the line in the first part of the description, and repeat the numbers in the condition notes. On eBay, make sure the measurements match what you enter in the item specifics for size. On Poshmark and Mercari, the description is the only place the buyer will look before making an offer.
Wording that works:
- Minor: "Tagged L, measures like a true M. Chest 20.5 in flat, length 24.5 in, sleeve 24 in. Slight shrinkage from washing. No holes, stains, or odor."
- Clear: "Tagged XL, has shrunk. Flat chest 21 in, length 23 in, sleeve 22.5 in. Brand's XL chart is 23 / 26 / 25.5. Fits like a M. Fabric intact, hem flares slightly."
- Significant: "Shrunk and felted. Flat measurements 19 / 21 / 20. Side seams twist forward. Priced as a small, soft, dense knit. Sold as-is."
Notice what these lines share: the tag size, the real numbers, the reference, and what is still fine. They never say "gently used" or "shrunk a bit." Specific beats soft every time. It also lets a buyer self-select, so the sweater goes to someone who wants that size.
Photograph the tape against the garment in the listing photos, along with the tag. That makes your tier visible without anyone having to take your word for it.
Putting the Grade Together
Run the five factors in order. Start with Fabric Condition (felting, hand, thinning), then Structural Integrity (twist, hem, collar), Cosmetic Appearance (pattern distortion, silhouette), Functional Elements (neck opening, cuffs), and Odor & Cleanliness. Then apply your shrinkage cap from the table. The final tier is the lowest cap that any factor, including measured shrinkage, justifies.
Consistency is the point. If your Good means 4% to 7% loss on Tuesday, it means that on Friday too. Buyers can't see your process, but they can feel it when your grades line up with what arrives in the mailbox. That is how you cut not-as-described cases on knits, and it is why one tape measure is worth more than any adjective.
Want to see how this plays out on a real garment? Grade one cotton sweater with GradeThread, add your flat measurements, and read the condition report across all five factors. One sweater is enough to tell you whether it fits the way you grade.