Three of the most expensive quality failures in streetwear share one thing: they only show up after the customer washes the garment.

By then, it has shipped, been sold, and worn.

The return is not a defect; you can’t hide behind an inspection photo; it is a photo the customer takes.

Shrinkage, spirality, and colorfastness are the three “wash failures.”

They are not obscure.

Every mill and every quality lab has standardized tests for them, and every serious buyer specifies limits against those tests.

The problem is that most first-time brand founders never write those limits down, so the factory defaults to whatever the greige fabric does, and the founder finds out in the customer’s washing machine.

This guide fixes that.

It covers what actually causes each failure, how the industry measures it, and the finishing processes that prevent it.

This part that protects you: the exact lines to put in your tech pack so the factory is accountable for the result.

If you can read a fabric spec, you can specify stability.

If you can’t yet, start with how to read a fabric spec sheet and the streetwear fabric guide, then come back.

The diagnostic · gibbenclothing

Shrink, Twist, Bleed

Every wash failure in a knit or woven garment is one of these three.

Learn to name which one you’re looking at because each has a different cause, a different test, and a different fix.

Confusing them is how founders “solve” the wrong problem.

Shrink:

Dimensional change

The garment gets smaller (or occasionally grows). A size M washes down to an S. Cause: tension locked into the fabric releases when water and heat hit it.

Twist:

Spirality/torque

The side seam rotates toward the front; the hem goes crooked; the print sits off-square. Cause: residual twist in the yarn unwinds in the knit loop.

Bleed:

Colorfastness loss

Dye migrates onto other garments in the wash, onto skin by rubbing, or fades under light. Cause: dye that was never properly fixed or washed off.

01 / Shrinkage:

What causes shrinkage and how to control it:

Fabric shrinks because fabric and yarns are under tension the entire time they are being made, and washing is the first chance they get to relax.

Yarn is spun under tension.

It is knitted or woven under tension.

It is stretched lengthwise through dyeing and finishing machines.

All of that stress gets locked into the structure.

When the finished garment meets water, heat, and mechanical agitation, the fabrics swell, the loops or crimp contract, and the fabric snaps back toward the relaxed state it “wanted” all along. That contraction is shrinkage.

Because the mechanism is stored tension, the direction that was pulled hardest shrinks the most.

In knits, that is usually the length (the wale direction).

In wovens, the warp yarns held taut on the loom typically shrink more than the weft.

This is why you rarely see a single shrinkage number that means anything.

You track length and width separately, and you never average them.

The three types you’ll actually meet:

Relaxation (residual) shrinkage is the big one: the fabric releasing manufacturing tension on the first few washes.

Good finishing removes most of it before the fabric ever reaches your cutting table, but “most” is not “all,” and the residual is what lands on your customer.

Progressive shrinkage is the natural-fiber tail: cotton and viscose can keep giving up a little more dimension over several washes, not just the first.

It is why a “3% fabric” can measure worse after five cycles than after one.

Felting shrinkage is a wool-specific mechanism (scales interlocking under heat and agitation) and mostly irrelevant to cotton-based streetwear but worth knowing so you don’t misdiagnose a blend.

How much is “normal”?

Commercial acceptance limits are well established, and they differ by construction because knits carry far more stored tension in their loops than wovens do in their crimp.

As industry benchmarks, wovens are generally held to around ±2-3%, and knits to around ±5%, with premium and workwear programs often tightening those further.

Raw fabric behavior sets the floor: polyester typically moves under 1%, cotton-rich jersey commonly wants to move 5-8% before finishing controls it, and viscose/rayon moves the most of all.

Typical unfinished shrinkage tendency by fabric (industry ranges verify against your own tested lots)
Fabric/construction Typical range* Notes for streetwear
Polyester (woven/knit) < 1% Thermoplastic; stable once heat-set
Cotton/poly blend (e.g., 65/35) 2–4% More poly, more stable
Cotton woven (poplin, twill) 3–5% Sanforizing pulls this well under 3%.
Cotton jersey/fleece (knit) 5–8% Compacting is the control; loose loops shrink most.
Viscose/rayon up to 8%+ Needs aggressive pre-treatment; watch for progressive shrink
*Ranges before finishing, after ~3 home laundry cycles. For product-specific fiber behavior, see the blend guide below.

The fabric choice underneath all of this is a separate decision with its own trade-offs covered in cotton vs polyester vs blends.

For stability specifically, the rule of thumb is simple:

The more natural fabric and the looser the construction, the more you must lean on finishing and pattern allowance to hold size.

How does the industry measure it?

You don’t argue about shrinkage; you test it.

The standard method is a controlled wash-and-dry cycle on a marked specimen, measured before and after.

The washing/drying procedure follows ISO 6330 (or AATCC 135/150 for automatic home laundering); specimens are prepared and marked per ISO 3759, and the dimensional change is calculated per ISO 5077.

The formula is unglamorous and exact:

Shrinkage formula:

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

Run it separately for length and width.

A negative result is growth, not shrinkage, and yes, knits can grow in width while shrinking in length.

Two operational points that separate a real test from a reassuring one.

First, test fabric from the bulk production lot, not a pristine lab dip.

The roll you’ll actually cut is the roll that matters.

Second, wash conditions decide the answer: temperature, agitation, and especially drying method.

Tumble drying pulls the most shrinkage, and line drying the least.

If your customers tumble dry and your test line dries, your test lied to you.

What a wash test actually involves:

You don’t need a lab to sanity-check a fabric, and knowing the steps lets you audit whether a supplier’s “test” was real.

A specimen is cut from the bulk roll, skipping the first few meters, where machine tension distorts the fabric, and marked with reference points a set distance apart in both length and width (knits are patted flat without stretching; edges are lightly stitched or overlocked so they don’t curl or fray).

The marked specimen is conditioned, washed, and dried according to the specified method; allowed to relax and recondition; then re-measured between the same marks.

Length and width are calculated and reported separately.

The whole point is repeatability: change the temperature, the load, or the conditioning time, and you change the number, which is why cross-border disputes almost always trace back to two parties running different conditions and each believing their own result.

How to control it before it reaches the customer?

There are two levers, and you want both.

The first is finishing compacting knits, sanforizing wovens, and heat-setting synthetics, which physically pre-releases the tension so it isn’t there to release later.

That is Section 4. The second is pattern allowance: if a fabric is going to shrink a known amount, the pattern is scaled up by that amount so the garment lands on-spec after washing.

If a body length must finish at 70 cm and the fabric shrinks 5% in length, you cut it at 70 ÷ 0.95 ≈ 73.7 cm.

Get the shrinkage number wrong, and every size is wrong in the same direction, which is why the wash test comes before the marker is made, not after.

02 / Spirality:

Spirality: why knits twist

Spirality, also called torque or twist.

It is the failure that surprises founders most.

The fabric can look perfectly flat and still betray you in the wash.

You notice it when the side seam of a tee has rotated toward the front, when the hem hangs crooked, or when a chest print ends up sitting at a slight angle.

In a knit, it is the courses (the horizontal loop rows) displacing from a line perpendicular to the fabric edge.

Measured as an angle of displacement, it is a rotation the garment develops as latent stress releases during laundering.

Where does it come from?

The root cause is residual twist in the yarn. Yarn is twisted during spinning, and a “lively” (over-twisted) single yarn wants to untwist.

In a single-jersey knit, the most common and most vulnerable streetwear construction, every loop is formed by one such yarn, and when the yarn relaxes, it rotates the loop, and thousands of rotating loops add up to a whole panel that twists.

Two design factors make it worse: loop length (longer, looser loops spiral more than short, tight ones) and yarn twist multiplier (more twist, more torque waiting to release).

Circular knitting machines can also bake in a baseline non-verticality from the direction the needles feed.

This is why spirality is overwhelmingly a knit problem and specifically a single-jersey problem.

Interlocks and ribs are more balanced by construction.

Wovens rarely spiral, though they have their own cousin skew and bow from uneven warp tension.

(Denim sometimes has spirality introduced on purpose for the leg-twist look, which is a good reminder that the goal is control, not zero.)

How is it measured?

Spirality is judged after laundering, not on the greige roll, because it is a released-stress phenomenon.

The standards are ISO 16322 (Part 2 covers woven and knitted fabrics; Part 3 covers garments) and the AATCC equivalent, AATCC 179.

A specimen is marked before washing and laundering, and the change in skew angle is measured.

It is reported as a percentage or an angle; many programs specify a maximum around 5% for single jerseys and tighter for premium.

Practically, what you demand is a spirality figure measured after the same wash cycle you use for shrinkage on the bulk lot, reported as a number you can pass or fail against, not a verbal “it’s fine.”

A supplier who tests shrinkage but shrugs at spirality is telling you they don’t control their single jersey.

On a tee program, that is the difference between a print that stays square and one that walks off-center by the tenth wash.

How to prevent it?

  • Balanced or plied yarn. Two-ply and low-torque/”S-Z” balanced yarns cancel much of the rotational stress before it reaches the loop. This is the most durable fix, and it starts at yarn selection, not finishing.
  • Sensible twist multiplier. Specifying a yarn with a lower twist factor for jersey reduces liveliness. It is a mill conversation, and it belongs in the fabric spec.
  • Compacting and heat-setting. Mechanical compacting and, for blends, heat-setting relax and lock the loop geometry, cutting the residual twist that would otherwise release in the wash. The same processes that fight shrinkage do double duty here.
  • Garment engineering. Where a fabric has a known small spiral, it can be partly designed around, but this is a patch, not a cure, and it doesn’t survive a spec audit. Fix it in the fabric.

03 / Colorfastness:

Colorfastness and dye-bleed risks:

Colorfastness is a fabric’s resistance to giving up its color by bleeding onto other fabrics in the wash, transferring onto skin or a sofa by rubbing, or fading under light.

Unlike shrinkage and spirality, which are mostly physics, color failures are mostly chemistry: dye that was never properly bonded to the fiber or excess unfixed dye that was never washed off after dyeing.

Deep shades (black, navy, and red) and reactive-dyed cottons are the usual suspects, and they are exactly the colors streetwear lives on.

The four failures and the four tests:

“Colorfastness” is not one number; it is a matrix.

You specify the ones that match how your garment is actually used.

The core four for apparel:

The colorfastness tests that matter for streetwear
Failure What it is ISO AATCC
Wash Color bleeds or stains other garments during laundering. 105-C06 61
Rubbing (crocking) Color transfers by friction, dry and wet. 105-X12 8
Light Color fades under sunlight / artificial light. 105-B02 16
Perspiration Sweat (acid/alkaline) causes fading or staining. 105-E04 15

Results are read against the grayscale, a 1-5 rating where 5 means no change or no staining and 1 means severe.

Two scales exist: one for color change in the specimen and one for staining onto an adjacent white fabric, and both matter.

A common commercial pass gate is Grade 4, with wet crocking often allowed a notch lower (Grade 3) because it is inherently harsher.

Dark and pigment-printed goods are where you watch crocking hardest.

Three things founders get wrong:

They specify a grade without a method. “Colorfastness Grade 4” is meaningless on its own.

AATCC 61 and ISO 105-C06 wash at different temperatures with different mechanics, so the same fabric scores differently. Name the test and the grade.

They mix standard families. Pick one family, ISO or AATCC, per purchase order and tech pack.

ISO is the default for multi-country export and AATCC for US-specific contracts.

Interchanging ISO 105-X12 and AATCC 8 on the same order is how you get disputes.

They set a grade the dye can’t hit. Asking for Grade 4 wash fastness on a cheap direct-dyed black is asking for a fail.

Ask the mill for the dye class first, then set a realistic gate or upgrade the dyeing (reactive with proper fixation and wash-off) to earn the grade you want.

The cheapest fix of all:

A huge share of wash-bleed is simply unfixed dye that was never rinsed out.

A proper after-wash / soaping step at the mill removes surplus dye before it can migrate.

If a supplier’s black is bleeding, the first question isn’t “what dye?” it’s “did you wash it off?”

Why the dye class matters:

Colorfastness starts with how the color got onto the cloth, and the three routes fail in different ways.

Reactive dyes form a chemical bond with cotton fiber and, when properly fixed and washed off, give strong wash fastness, but they depend entirely on that fixation-and-wash-off being done right, which is where cheap dyeing cuts corners.

Pigment (including most water-based and plastisol prints and pigment-dyed/”garment-dyed” looks) sits on the fabric surface, held by a binder rather than bonding into it, so it tends to give softer, faded aesthetics and is more prone to crocking (surface color rubbing off, especially before it’s fully cured).

Direct dyes are cheap and simple and generally the weakest on wash fastness, which is why a bargain black tee is often the one that stains the wash.

You don’t need to run the dyehouse, but you should ask which class your color uses, because it sets the ceiling on the fastness grade you can realistically demand.

Where garments are dyed or washed after sewing, garment dye, stone wash, enzyme color, and shrinkage behavior all shift, so those processes get tested as finished garments, not just piece goods.

That whole decision set is its own topic; see garment dyeing and washes.

04 / Finishing:

Pre-shrinking & finishing solutions:

Almost everything in the first three sections is solved in the same place: the finishing line.

Finishing is where the stored tension that causes shrinkage and spirality gets pre-released and where dye gets fixed and washed off.

A brand that understands finishing can look at a factory and know, before a single garment ships, whether its stability is engineered or accidental.

Here are the processes to know and how to verify each one exists.

Compacting for knits:

A mechanical process that uses steam and pressure to compress the knitted loops, physically pre-releasing relaxation shrinkage and tightening the loop geometry.

This is the control for jersey and fleece: it is what takes a cotton knit from a wild 6-8% down toward a shippable range, and it cuts spirality at the same time.

If a knit-based brand’s factory can’t tell you about its compactor, that is a red flag.

Sanforizing (compressive shrinkage) for wovens:

The woven-cotton equivalent.

Fabric is fed through a rubber belt that mechanically compresses the warp yarns into a pre-shrunk, stable state.

A properly sanforized woven cotton holds well under 3%.

A weave that skips or under-runs sanforizing shows strong warp shrinkage, the classic “the waistband got tight after one wash” complaint on bottoms.

Heat-setting for synthetics and spandex blends:

Polyester and elastane are thermoplastic: heat can set them into a stable shape that then resists further movement.

Essential for any spandex blend (activewear, stretch bottoms) and for stabilizing poly-rich knits.

Skipped heat setting is a common cause of blend garments that twist or grow unpredictably.

Wet processes: garment wash, pre-wash, soaping

A pre-wash or garment wash relaxes fabric in water before the customer ever does, so much of the residual shrinkage happens on the factory floor instead of at home.

The soaping / after-wash step in dyeing is what removes unfixed dye and buys you colorfastness.

These are cheap insurance that many low-cost suppliers quietly cut to save time and water, which is exactly why they show up as failures downstream.

Drying the quiet variable:

How fabric is dried after wet processing matters as much as the wash.

Relaxation drying (tumble or vibration/”through-air” drying that lets the fabric relax with minimal tension) preserves the stability the earlier steps bought.

Stretching fabric back out on a tenter frame under length tension to hit a width target can quietly reintroduce the very shrinkage you paid to remove.

Trade-off to know:

Softeners and silicone finishes make fabric feel premium, and buyers love them, but heavy softeners can reduce crocking fastness and mask a poorly washed-off dye.

A gorgeous hand and a marginal wet-crock number sometimes come from the same bottle. Feel and fastness are a balance, not a free lunch.

The audit takeaway: You don’t need to run the machines, but you should be able to ask, for your specific fabric, which of these steps it goes through and see the resulting test numbers.

“We finish it properly” is not an answer. “It’s compacted, garment-washed, and here’s the after-wash shrinkage and spirality report on your lot” is.

05 / The tech pack:

Specifying limits in the tech pack:

Everything above is knowledge.

This is the part that protects you.

A factory is accountable for the numbers you write down and nothing else.

If your tech pack says nothing about stability, the factory owes you nothing about stability, and a “the shirt shrank” complaint after delivery is your problem, not theirs.

Written tolerances are what turn a wash failure from a dispute into a clear pass/fail the supplier agreed to in advance.

So you write four things.

We call it “The Stability Spec,” the block that belongs on every tech pack, right next to the measurements.

(For where it sits in the wider document, see how to write a tech pack.)

The Stability Spec: the four lines every tech pack needs

01

Max shrinkage, length & width are stated separately. e.g., Length ≤ 5%, Width ≤ 5% (knit) or ≤ 3% (woven). Name the wash method and conditions.

02

Max spirality, e.g., ≤ 5% per ISO 16322 / AATCC 179, especially non-negotiable for single-jersey tees.

03

Colorfastness minimums, by test and grade, one standard family, e.g. Wash (ISO 105-C06) ≥ 4 · Dry crock (X12) ≥ 4 · Wet crock ≥ 3 · Light (B02) ≥ 4.

04

Which lot gets tested, and the remedy if it fails, e.g. Test bulk production lot, not lab dip. Fail = re-finish or reject at supplier cost.

Four lines.

That is the entire difference between a founder who eats returns and one who doesn’t.

A few notes on getting them right:

Set gates where the fabric and dye can actually meet:

A tolerance you specify, but the fabric can’t hit; it just guarantees a fight at inspection.

Before you lock a number, ask the mill what the fabric and dye class realistically deliver, then either accept a sensible gate or upgrade the fabric/finishing to earn a tighter one.

A realistic Grade 4 you enforce beats an aspirational Grade 4-5 you wave through.

Match the test to real use, then hold to one family:

Add perspiration fastness for anything sweat-facing; lean on crocking for dark and pigment-printed goods; and add light fastness for anything that lives outdoors.

But keep the whole PO in one standard family (ISO or AATCC) so procedures don’t collide.

Test the lot you’re shipping:

Every number above should be run on the bulk production fabric, not a perfect early sample.

Tie your acceptance to it, alongside your general defect standard. Your defect AQL is a separate acceptance framework layered on top. It governs visible defects; the Stability Spec governs performance.)

consistency note:

Whatever wash conditions and grade gates you standardize here should match what your sampling, tech-pack, and QC guides state, so a founder reading across the cluster sees one house standard, not three. Reconcile the exact numbers once and reuse them.

The pre-production stability checklist:

Before bulk is cut, walk through this once.

It is the short version of everything above, in the order the decisions actually happen:

  1. Wash-test the bulk lot length and width shrinkage, plus spirality, on the actual production fabric under your customer’s wash conditions. Not the lab dip.
  2. Confirm the finishing. Ask which processes your fabric goes through (compacting/sanforizing/heat-setting/garment wash) and see the resulting numbers, not a reassurance.
  3. Set the pattern allowance from the tested shrinkage, length, and width separately before the marker is made.
  4. Lock colorfastness gates by test and grade, one standard family, matched to the dye class the mill confirmed.
  5. Write the Stability Spec into the tech pack with the remedy clause, and reference it in the manufacturing agreement.
  6. Tie acceptance to it alongside your defect AQL so that a lot that fails performance is caught at the same gate as a lot with visible defects.

Common questions

Yes, with natural fabric, effectively always.

The goal isn’t zero; it’s controlled and known.

Good finishing pushes residual shrinkage into a small, predictable range, and pattern allowance absorbs what’s left so the garment lands on-spec after washing.

A “pre-shrunk” fabric has had most, not all, of it removed.

Almost always because the sample and bulk weren’t the same fabric lot or weren’t tested the same way.

Early samples are often made from carefully finished piece goods; bulk can come from a lot that was rushed or under-finished.

Always run the wash test on the actual bulk lot under the wash conditions your customers use.

Shrinkage changes the garment’s size; spirality changes its shape/alignment by rotating it.

A garment can pass shrinkage limits and still fail on spirality; the side seam twists to the front even though the measurements are fine.

They’re tested separately and fixed differently, though compacting helps both.

For most streetwear: always wash and crock (dry + wet); light fastness if the piece lives outdoors; perspiration if it’s sweat-facing.

Dark shades and pigment prints need crocking, which is watched the hardest. Specify each by test method and grade, and keep everything in one standard family (ISO or AATCC).

Most often, it’s unfixed dye that was never washed off after dyeing, not the dye itself.

A proper after-wash/soaping step removes surplus dye before it can migrate.

Ask the supplier about their wash-off process and retest wash fastness (ISO 105-C06 / AATCC 61) on the bulk lot before shipping more.

Pattern allowance is half the answer, not all of it.

It compensates for a known, consistent shrinkage, but if the fabric wasn’t finished properly, shrinkage will be high and inconsistent lot-to-lot, and no allowance can chase a moving target.

Fix stability in finishing first, then allow for the predictable remainder.

Proper finishing and dye wash-off do carry cost, and a supplier competing purely on price may have cut exactly those steps.

But the alternative cost is return rates, replacement units, and reputation.

Specifying stability doesn’t create the cost; it moves it from your customers’ washing machines to the quote, where you can see it.

Whoever the tech pack says it is.

If you specified tested limits and a remedy clause, a failing lot is the supplier’s cost.

If you specified nothing, it’s yours.

This is the entire reason the Stability Spec exists; it assigns accountability before production, in writing.

(This is general guidance; put actual remedy terms in your manufacturing agreement and have it reviewed.)

Reference:

Fabric-risk glossary:

Relaxation (residual) shrinkage:

Dimensional change from the fabric releasing tension stored during spinning, knitting/weaving, dyeing, and finishing is the main shrinkage type in apparel.

Progressive shrinkage:

Additional small shrinkage that natural fibers give up over several washes, not just the first cycle.

Spirality (torque/twist):

Rotation of knit courses/wales from perpendicular, released during laundering, shows as twisted seams and crooked hems.

Twist multiplier / lively yarn:

A measure of how much twist is in a yarn; over-twisted (“lively”) singles drive spirality in single jersey.

Crocking:

Color transferring by rubbing/friction was tested dry and wet; the wet result is usually the weaker of the two.

Grey scale:

The 1-5 rating is for color change and for staining; 5 is best. Grade 4 is a common commercial pass gate.

Compacting:

Steam-and-pressure process that pre-releases relaxation shrinkage and tightens loop geometry in knits.

Sanforizing:

Compressive pre-shrinking of woven cotton via a rubber belt; the woven equivalent of compacting.

Heat-setting:

Using heat to stabilize thermoplastic fibers (polyester, elastane) so they resist further dimensional change.

Wash-off / soaping:

The dyeing step that removes unfixed surplus dye; skipping or shortcutting it is a leading cause of wash bleed.

Pattern (shrinkage) allowance:

Scaling the cutting pattern up by the expected shrinkage so the garment finishes on-spec after washing.

Grey/greige goods:

Fabric straight off the machine, before finishing, was stored with the highest tension and shrinkage tendency greatest.

From the manual

This is one chapter of The Complete Streetwear Manufacturing Guide

These chapters cover taking a streetwear brand from idea to shipped units, costing, sourcing, tech packs, QC, freight, and launch. All written from the Sialkot factory floor.

Open The Full Guide

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.