You ship a clean tee. The fit is right. The customer washes it once.

Now the body is short, the side seam has crawled toward the front, and the hem waves.

Nothing tore. Nothing stretched. The shirt just refuses to hang straight.

This is the most common quality complaint on jersey tees. It is almost never bad luck.

It is a finishing decision that was skipped or a spec you never wrote.

Here is the thing a hoodie brand never has to think about.

A hoodie is fleece plus a second layer; it hides movement.

A tee is one thin layer. Call it the single-layer penalty: with nothing to hide behind, a 5% length loss and a rotated seam read instantly on the body.

The same defect that a hoodie absorbs, a tee wears on its face.

So a tee’s tolerances have to be tighter than a heavier garment’s, not the same.

That penalty runs through every control below.

This is the production-side list what you demand from the mill and the sample room, in the order it fails.

Different garment:

Making hoodies, not tees? The fleece failure mode is different; see Why hoodies shrink unevenly. For the underlying fabric science across both, the canonical reference is our Shrinkage, Spirality & Colorfastness guide.

SHR-01

What actually causes tee shrinkage and twisting?

Two different problems get blamed on the same word.

Shrinkage is the garment getting smaller. Twisting (spirality) is the garment rotating out of square.

A tee can do one, the other, or both. They have different causes, so treat them separately.

The cause of shrinkage: released tension

A knit tee is under stress the whole way through production.

Yarn is spun under tension. The machine holds the loops under tension.

Dyeing, drying, and finishing stretch it again. Every step leaves the yarn longer and thinner than it “wants” to be.

That stored stress does not disappear. It waits.

When the garment hits water, heat, and agitation, the fibers swell and the yarns finally relax back toward their natural, shorter, fatter shape. The fabric contracts.

Every knitted fabric carries residual internal stress from spinning, knitting, dyeing, and finishing and releases it toward a relaxed state, which is why the first wash does most of the damage.

Two types worth naming:

  • Relaxation shrinks: the big one. Happens almost entirely on the first wash as stored tension releases. This number decides whether your fit survives.
  • Progressive shrinkage: a smaller amount that keeps accumulating over repeated washes, common in cotton- and viscose-rich fabrics. This is why serious programs test over multiple cycles, not one.

Three production factors decide how bad relaxation shrinkage gets:

  1. Unrelieved knit tension: fabric that was never allowed to relax before it was cut.
  2. No compacting: the mechanical pre-shrinking step (SHR-02) was skipped or run soft.
  3. Heat: hot wash and, worse, hot tumble dry drive the fibers to relax harder and faster.

Fabric weight matters too. A heavier, tighter jersey with a higher GSM hides less movement than a thin, open jersey.

If your tee is light and airy, your shrinkage tolerance has to be tighter, not looser. For how weight is measured and specified, see our GSM guide.

The cause of twisting: live torque, briefly:

Twisting is a different animal. It comes from the yarn: ring-spun singles are twisted during spinning, that twist stays “live,” and single jersey and an unbalanced knit have nothing to cancel it.

Finishing freezes the loops flat; the wash frees them; they rotate; the body turns.

Retailers cap the rotation at 5% or less, and it shows up in the first three to five washes. It’s a raw-material and knitting fault, not a sewing one.

Go deeper:

The full mechanics twist liveliness, the 0.73 twist-factor balance ratio, feeder geometry, and S/Z compensation live in the canonical Shrinkage, Spirality & Colorfastness guide. Below is only what changes the shape of a tee.

Where the single-layer penalty bites: three tee-only failure modes?

Because a tee hides nothing, three failures that a hoodie shrugs off are the ones that get a tee returned.

1. Tubular vs. side seam. A seamless (tubular) tee is knitted in the round, so its residual torque has no seam to anchor it; the whole tube rotates, and the fabric spirals around the body.

A side-seamed tee at least pins the rotation at two seams.

If you sell tubular blanks, your spirality tolerance has to be tighter and your yarn balance non-negotiable, because there is no seam doing you a favor.

2. The rib collar. The neck rib is a different fabric, knitted tighter, on a different machine, and it relaxes on its own schedule.

When the body shrinks more than the collar, the neckline ripples or stands up wavy; when the collar over-relaxes, it goes slack and won’t recover.

This is the tee’s signature “looks cheap after one wash” failure, and it is a collar-to-body mismatch, so it has to be tested and spec’d as its own line (SHR-05, SHR-06), not folded into the body number.

3. Print vs. fabric. The decoration and the jersey shrink at different rates.

A thick plastisol screen print (PRT-01) sits on top of the fabric and barely moves; when the jersey shrinks underneath it, the ink cracks, bubbles, or lifts into “islands.”

DTG (PRT-02) and water-based inks penetrate the fabric and move with it, so they crack far less.

If your tee is a heavy-print graphic tee, the print area is a shrinkage zone of its own; wash-test the printed panel, not a blank swatch.

Quick diagnosis: which tee failure do you have?

QC decision aid · tee symptom → cause → control
What you see after one wash The real problem Where it’s fixed
The collar ripples, stands wavy, or goes slack. Collar-to-body shrink mismatch SHR-05 / SHR-06
Print cracks, bubbles, or lifts into islands. Ink-vs-jersey differential shrink PRT-01 / SHR-06
Whole body spirals; seam crawls (worst on tubular). Spirality / live torque SHR-01 / SHR-02
Body & sleeves shorter. Relaxation shrinkage (length) SHR-02 / SHR-05
Fits tighter across the chest, no rotation. Width shrinkage SHR-02
Smaller every wash, not just the first. Progressive shrinkage SHR-03 / SHR-04

The takeaway: shrinkage is a finishing problem (tension you failed to relax); twisting is a yarn and knitting problem (torque you failed to balance); and on a tee, both get magnified by the single-layer penalty and by the collar and the print. The fabric choices behind all of it sit in our streetwear fabric guide.

SHR-02

Pre-shrinking and compacting: the finish to insist on

If you demand one thing from your mill, demand this.

Compacting is the mechanical step that pre-shrinks knit fabric on purpose, in the factory, so the customer’s wash has far less left to release.

This is the key spec. A tee that skips it is a tee that shrinks on the buyer’s first wash.

What to demand without the machine tour?

Compacting gathers the fabric instead of stretching it:

The fabric is overfed (fed in faster than it comes out), then steam and heated felt set it in that shortened, relaxed state so it leaves the line flat and stable.

On a tee, that’s the difference between a first-wash return and a garment that holds.

The one number worth knowing: shrinkage control peaks around 70-75% overfeed, so under-run compaction is a real and common shortcut.

What you actually put in the spec is short: fabric must be compacted, tubular, or open-width, with relaxation drying first.

Tubular compaction reaches the lowest residual shrinkage and also helps spirality worth insisting on for tubular tee blanks.

Go deeper:

Overfeed percentages, felt tension, steam-box relaxation, and relaxation drying, plus the tubular-vs-open trade-off in full, are in the Shrinkage, Spirality & Colorfastness guide.

“Sanforized” vs. “pre-shrunk”: don’t get played

Sanforized® is a licensed trademark with a real tolerance behind it:

Garments held to it shrink no more than ≈1% woven and ≈5% knit under standard washing, and only licensed mills can use the mark.

“Pre-shrunk” is generic and unregulated; many “pre-shrunk” garments got some compaction but may never have been held to that tolerance.

The phrase promises nothing you can enforce.

So do not accept “pre-shrunk” on a spec sheet. Accept a number of a maximum shrinkage percentage against a named test (SHR-08). Language is not a control. A tolerance is.

Why does some shrinkage always remain?

Even perfect compacting leaves a little. Compaction is mechanical, not chemical:

It shortens the cellulose, but the hydrogen bonds reform on the first home wash in a slightly tighter arrangement than the factory achieved.

Most residual shows up by wash one and is basically gone by wash three, after which a well-finished tee is stable.

The goal is never “zero.” It is residual shrinkage inside tolerance, plus a pattern cut to absorb what’s left (SHR-05).

SHR-03

Cotton vs. blend shrinkage behavior:

Fiber choice sets your risk ceiling before finishing even starts. Pick the fabric, and you have mostly picked the problem.

100% cotton jersey: the high-shrink, high-twist case

Cotton loves water. The fabric swells, the yarns relax, and knit cotton moves.

Untreated cotton jersey commonly loses up to about 5% in length and ~3% in width, and without any pre-shrinking, cotton can run 5–8%.

Sanforizing pulls that to roughly 1-2%.

Real-world data is messier than the label.

Across “pre-shrunk” ring-spun cotton tees, measured shrinkage has ranged from essentially 0% up to 5.33% under hot wash and hot dry even on garments marketed as pre-shrunk.

The lesson: cotton’s finish and wash conditions decide the outcome as much as the fabric.

Cotton is also the worst offender for twisting, carrying live torque with nothing to balance it.

Still the right call for a premium hand and print; you just have to control it, hard.

Polyester and poly-rich blends: the low-shrink case

Polyester barely absorbs water, so it barely relaxes.

Polyester knits typically shrink under ~0.5-2%, even after several washes.

A poly component acts like a skeleton, restraining the cotton.

Move to a 50/50 or 60/40 blend, and both shrinkage and spirality drop.

The trade-off is hand and print feel a brand decision, not a technical one, but know you are buying dimensional stability with it.

The practical rule:

  • 100% cotton single jersey → highest shrink, highest twist. Non-negotiable compacting + tight tolerance + wash-tested sample.
  • Cotton/poly blend → materially lower risk, at some cost to hand.
  • Higher GSM / tighter knit → less twist, easier stability than thin, open jersey.

Match the fiber to how the customer will actually launder the tee.

A 100% cotton streetwear tee sold to buyers who hot-wash and tumble dry is a return waiting to happen unless finishing and pattern absorb it.

Full fiber-by-fiber trade-offs are in our streetwear fabric guide.

SHR-04

Reading a shrinkage test report:

A test report is your accountability document. If you cannot read it, you cannot hold anyone to it.

Here’s how to read one in five minutes.

What to check in five minutes?

A report worth trusting names its method up top: AATCC 135 or ISO 6330 for size, AATCC 179 or ISO 16322 for twist.

Name none of these, and it isn’t a controlled test; reject it.

And note the method only measures the change; it sets no pass/fail line. Your spec does.

Go deeper:

The AATCC-vs-ISO recipe differences, machine settings, and method history are covered once in the Shrinkage, Spirality & Colorfastness guide. Here it’s only what you do with the numbers on a tee.

The four fields to read:

  1. Length vs. width, separately. Length (wale) shrinkage is the bigger problem on tees; one blended number hides it.
  2. Cycle count. The single-wash figure is the fit-killer; it captures relaxation shrinkage. Multi-cycle exposes progressive shrinkage.
  3. Wash temp & dry method. A 30°C/flat-dry and a 60°C/tumble-dry result are different universes, and they must match your care label and your bulk test. Lock one recipe across sample and bulk, or you’re comparing two experiments.
  4. Conditioning. Legit labs rest specimens (commonly ≥4h) before measuring.If that’s missing, the numbers drift.

Formula:

Shrinkage % = ((Original − Final) ÷ Original) × 100

The tee benchmarks:

No universal pass number; your spec is the authority.

For a knit tee, the working caps are ≤5% length (some programs allow 5-8% by end use), ≤3% width, and ≤5% torque/skew.

Sanforized® knit holds to ≤5%. The full woven/knit/blend tolerance table is in the hub guide.

A worked read: the “watch it” pass

Say the report shows, on a single 40°C wash, tumble-dry low: length −4.6%, width −2.1%, skew 3%.

Against a spec of length ≤5%, width ≤3%, and skew ≤5%, this passes but only just on length.

Two takeaways: the pattern needs a ~4.6% length allowance (SHR-05), and there’s little headroom, so any drift in bulk fabric could push it over.

A result at −4.6% is a “watch it” pass, not a comfortable one.

A tolerance you cannot see on paper is a tolerance you cannot enforce.

Get the report, read the length and width, check wash conditions, compare to spec.

SHR-05

Designing the pattern for residual shrinkage:

Compacting reduces residual shrinkage.

It does not delete it. So the pattern has to absorb what’s left.

This is the step brands forget and then blame the mill for a fit that was never patterned to survive a wash.

Cut oversized, wash down to target:

Pattern the garment slightly larger than the finished spec so after washing it nets down to the size you want.

Work backwards from tested numbers:

  1. Take the measured residual shrinkage from your wash-tested sample (SHR-06), say 4% length and 2% width.
  2. Add that as an allowance to the block, in each direction, so the post-wash garment lands on target.
  3. Working adjustment: to net 70 cm body length after 4% shrinkage, cut ~73 cm (target ÷ (1 − shrinkage): 70 ÷ 0.96 ≈ 72.9 cm).

Never guess the allowance. Base it on the actual tested percentage for this fabric and finish two “100% cotton jersey” fabrics can behave differently.

Worked example · both directions

Target finished tee: 71 cm length, 52 cm chest (half). Tested residual: 4.6% L, 2.1% W.
Cut length = 71 ÷ (1 − 0.046) = 71 ÷ 0.954 ≈ 74.4 cm
Cut width = 52 ÷ (1 − 0.021) = 52 ÷ 0.979 ≈ 53.1 cm

Add those allowances, wash-test the next sample, and confirm it nets to 71 × 52.

If it overshoots or undershoots, the residual number was off; adjust and retest.

The pattern is only “right” once a washed garment lands on the finished spec.

Allow for skew, too:

Shrinkage allowance protects size. It does not protect shape.

If the fabric has residual spirality, no length allowance fixes a rotating seam.

Two defenses:

  • Fix it upstream with balanced/plied yarn, correct S/Z twist compensation, and proper compaction (SHR-01–02). Prevention beats rework.
  • Pattern with skew awareness: some mills pre-set a counter-skew so fabric relaxes to square. Test it, because forced skew-setting can partly revert after repeated washes.

Let the fabric rest before cutting:

One free, high-impact control: relaxation time.

Fabric pulled straight off the roll is still under tension; cut it now and it moves after sewing.

Let the rolls relax, commonly 12-24 hours before spreading and cutting, so they reach a true relaxed state first.

Skip the rest and you cut panels that are still “long,” which then shrink after they’re already sewn. A self-inflicted fit failure.

Measure at defined points.

Allowances only work if everyone measures the same landmarks a fixed distance below the HPS for length and a fixed distance below the armhole for chest.

Vague points produce vague allowances.

SHR-06

Wash-testing the sample before bulk:

This is the cheapest insurance in the whole process.

One controlled wash on the pre-production sample tells you what 3,000 units will do.

Catch it here, and bulk is safe. Skip it and you find out after the containers ship.

Test the garment, not just the fabric:

A fabric swatch and a finished tee shrink differently.

The finished garment has seams, neck tape, hems, and trims, and those interact.

So wash-test the actual sewn sample, made from bulk fabric on bulk machines.

It’s the only test that answers the real question: does this product hold its size and shape?

The protocol:

  1. Mark reference points before washing a fixed body length, a fixed chest width, a straight line down the side seam, and the tee-specific ones the collar width and depth, plus a fixed span across any printed panel. The collar and the print move on their own; measure them or you won’t catch the two failures buyers notice most.
  2. Wash and dry to a defined program, ideally the harshest your care label permits (SHR-07), using AATCC 135 or ISO 6330 parameters.
  3. Single wash first captures the make-or-break relaxation shrinkage.
  4. Then 3-5 cycles expose progressive shrinkage and let twisting fully develop, since torque works out over the first several washes.
  5. Re-measure. Compute shrinkage % per direction, and measure how far the side-seam line moved for skew.

Read the result against spec:

  • Inside tolerance in every direction, seam movement under ~5%? Approve bulk.
  • Out of tolerance? Do not approve. Fix finishing (overfeed/compaction), fiber, or pattern allowance, then retest.

Correcting spirality after bulk usually means costly alteration or full rework; minor cases can sometimes be re-compacted or pressed, but severe twist is a scrap-or-rework event.

The wash test keeps you out of that hole. For the full sample-approval workflow, this sits inside; see our clothing sampling guide.

SHR-07

Care labels that protect fit:

You’ve controlled the fabric, the finish, and the pattern.

The last variable is the customer’s laundry, and the care label is the one place you get to steer it.

The dryer does most of the damage.

The same cotton tee that loses 2-5% hot-washed and tumble-dried drops under 1.5% cold-washed and air-dried, because heat plus agitation is exactly what releases stored tension and live torque.

So the label is a fit-protection tool, and on a tee it has two extra jobs the hoodie label doesn’t: protect the collar and protect the print.

The tee-tuned instruction set:

  • Wash cold, inside out gentler on fit, and it puts the print on the inside, away from abrasion.
  • Don’t tumble dry (or tumble low) the single highest-leverage line and the one that saves the print from cracking.
  • Reshape the collar and dry flat. A wet rib collar sets in whatever shape it dries in; flat-drying is how it settles square instead of wavy.
  • Don’t iron the print.

Keep it honest: care instructions are a trust signal to buyers and to Google’s raters, so don’t print “tumble dry” if your own bench test (SHR-06) fails at that setting; match the label to what the tee actually survived.

Full symbols, wording, and legal requirements live in the care label guide.

SHR-08

Specifying shrinkage limits: the enforceable standard

Everything above collapses into one deliverable:

A written, testable shrinkage and shape spec in your tech pack.

Without it, every control here is a favor your supplier may or may not do.

With it, they’re contractual.

Why is the spec the real control?

“Please make sure it doesn’t shrink” is not enforceable.

A tech pack stating a maximum percentage against a named test, at a named temperature, is.

When a bulk comes back over tolerance without this line, there’s no basis to reject or charge back. The spec is the leverage.

Go deeper:

The general principle of writing dimensional-stability limits into a tech pack is covered for all fabrics in the Shrinkage, Spirality & Colorfastness guide. Below is the block cut for a tee with the collar and print lines a fabric-level spec won’t have.

What does a complete spec include?

  1. Shrinkage tolerance, per direction. e.g., ≤5% length, ≤3% width. Both numbers. Length is where tees fail.
  2. Torque / skew tolerance, e.g., ≤5% seam displacement.
  3. Named test method. AATCC 135 / ISO 6330 for size; AATCC 179 / ISO 16322 for twist.
  4. Wash conditions. Temperature, cycle, and dry method matched to your care label.
  5. Cycle count. Single wash for the pass/fail relaxation number, plus a multi-cycle check.
  6. Finishing requirement. Fabric must be compacted; if required, held to Sanforized® tolerance, not merely “pre-shrunk.”
  7. Test on the finished garment, from bulk fabric, not just a swatch.

Copy-paste spec block:

Shrinkage & shape control spec:

Fabric: [single jersey / weight GSM / fiber]

Finishing: Compacted (tubular or open-width) REQUIRED.            

Held to Sanforized tolerance: [Yes / No].

Test size: AATCC 135 (or ISO 6330), [40°C wash, [tumble low / flat] dry].

Test twist: AATCC 179 (or ISO 16322).

Cycles: 1 wash (pass/fail) + [5] cycle check.

Tolerance: Length ≤[5]% | Width ≤[3]% | Skew/torque ≤[5]%.

Collar: Rib recovers to spec; no wave/slack after [5]-cycle wash.

Print: No crack/lift on printed panel after wash (test printed unit).

Test unit: Finished garment, bulk fabric, bulk machines.

Approval: A wash-tested pre-production sample must pass before bulk.

Make it a required field on every tee tech pack.

A supplier who cannot hit a written tolerance tells you that before you place the order, which is the entire point.

Supplier excuses, answered:

When a bulk fails and the reasons start, here’s the terse reply to each:

“Cotton always shrinks.” True, which is why the spec allows a tolerance.

This lot exceeded it. Show me the compacting overfeed and the wash-test report.

“It’s pre-shrunk.” “Pre-shrunk” is not a number.

The spec says ≤5% length against AATCC 135. Where is the result?

“The sample was fine.” Then the bulk fabric or finishing drifted.

Test the bulk fabric and finished garment with the same wash recipe.

“It’s a sewing issue.” Seam rotation from spirality is a yarn and knitting fault, not a sewing one. Show the yarn twist and balance.

“Customers wash it hot.” The care label says cold / low.

We tested at the label’s harshest allowed setting, and it still must pass.

None of these are winnable without a written spec and a wash-tested sample.

That’s why SHR-06 and SHR-08 exist.

Frequently asked

The single-layer penalty: a thin jersey has less mass and a more unbalanced structure than heavy fleece, so shrinkage and twist show up far more visibly.

A hoodie hides the same movement in its loft.

Hoodies have their own, different failure (the rib-vs-body differential): see why hoodies shrink unevenly.

The ink and the jersey shrink at different rates.

A thick plastisol screen print sits on top of the fabric and barely moves, so when the jersey shrinks underneath it, the ink cracks or lifts into “islands.”

DTG and water-based inks penetrate the fabric and move with it, so they crack far less. Wash-test the printed panel, not a blank swatch.

No. Sanforized® is a licensed trademark with an enforceable tolerance (≈≤5% knit).

“Pre-shrunk” is generic and unregulated; it may mean some compaction or very little. Specify a number, not a word.

Not in cotton. Compaction is mechanical, so a small residual remains and mostly appears by the first wash, settling by about the third.

The goal is residual shrinkage inside tolerance, plus a pattern that absorbs it not zero.

That’s spirality/torque: live yarn twist in an unbalanced single-jersey structure rotates as the fibers relax in the wash.

Balanced/plied yarn, correct twist direction, and proper compaction control it.

A single-wash AATCC 135 (or ISO 6330) on the finished pre-production garment, at the harshest care-label conditions.

That single number captures the relaxation shrinkage that decides fit.

Yes. The same cotton tee can shrink 2-5% hot-washed and tumble-dried versus under 1.5% cold-washed and air-dried.

Tumble-dryer heat is the biggest single driver.

From the T-shirt manual

This is one chapter of The Complete T-Shirt Manufacturing Guides:

These chapters cover taking a T-shirt brand from idea to shipped units, costing, sourcing, tech packs, QC, freight, and launch comparisons. All written from the Sialkot factory floor of Gibben Clothing.

Open The Full T-Shirt Guides

On the floor · Sialkot

Written by

Faizan Ahmad

Chief Apparel Technologist & Head of Manufacturing, Gibben Clothing · Sialkot, Pakistan

Faizan leads production at Gibben Clothing, a cut-and-sew streetwear manufacturer in Sialkot, with 8+ years turning raw yarn into retail-ready hoodies, tees, bottoms, jackets, tracksuits, and headwear. He doesn’t just write about clothing; he works the floor, so every guide here is grounded in real fabric behavior, QC standards, and production data from live runs.