Knitwear Lab Testing and AQL Inspection: Which Tests to Specify and Why

An AQL inspection is the last cheap decision in a knitwear order. It happens before the goods leave the factory, when a problem costs a rework or a conversation rather than a container of unsellable stock. Yet most buyers treat it as a formality and never look at what is actually being tested — or what is not.

There are two distinct quality systems, and conflating them is the most common mistake in a quality discussion. Laboratory testing measures how the fabric performs: will it shrink, pill, fade or contain a restricted substance. Inspection measures whether the finished garments match the approved specification. You need both, and one cannot substitute for the other.

The Knitwear Test Matrix

Every test below answers a specific commercial question. Specify a test because you need the answer, not because a checklist mentioned it.

Two pairings cause most of the confusion in practice.

ISO 6330 and ISO 5077 are used together, and neither works alone. ISO 6330 defines the domestic washing and drying procedure — the machine type, cycle, temperature and drying method. ISO 5077 defines how the dimensional change is measured. A result quoted as "shrinkage tested" without stating which wash procedure was used is not a result you can act on, because the same sweater shrinks differently on a delicate cycle and a hot wash with tumble drying. Specify both: the procedure *and* the measurement standard.

For pilling, check which part of ISO 12945 is being applied. The standard has four parts, all in the 2020 edition: Part 1 is the pilling box method, Part 2 is the modified Martindale method, Part 3 is the random tumble method, and Part 4 covers assessment of pilling, fuzzing and matting by visual analysis. In knitwear, ISO 12945-2 is the common choice, and many purchase orders pair it with Part 4 for the rating. Note that on the US side, ASTM D4970 is a Martindale-based pilling method with similar logic but its own scale, so the two are not directly interconvertible.

Understanding AQL and Defect Classification

Inspection sampling follows the AQL method — acceptance quality limit — the methodology standardised in the ISO 2859 series, with the US counterpart tables in ANSI/ASQ Z1.4. The 2026 edition of ISO 2859-1 replaced the long-standing 1999 version, so if your PO names a version, name the current one.

AQL works by accepting a statistically defined defect level rather than demanding perfection. The sample size for a given lot comes from a table indexed by lot size and inspection level; you then compare the number of defects found against the acceptance and rejection thresholds.

Defects are classified in three tiers rather than counted together:

AQL 2.5 is the most widely used level for general consumer goods; AQL 1.0 is applied where risk is higher, and AQL 4.0 where minor cosmetic variation is acceptable. The important nuance is the split: mixing critical defects into a 2.5 calculation understates the risk, because a safety failure should never be accepted at any rate.

To see how the arithmetic works, consider a 4,000-piece order inspected at General Inspection Level II. That lot size maps to code letter L, which samples 200 pieces. At AQL 2.5, the acceptance number is 10 — up to 10 defective pieces passes, 11 rejects the lot. Buyers who have never looked at the numbers are often surprised by how much variation sits inside an "acceptable" result.

What an Inspection Actually Checks

A knitwear inspection examines the garment against the approved specification, and the checklist clusters around a few themes:

• Measurement against the approved size spec, on the flat, with tolerance
• Workmanship — knitting faults, dropped stitches, linking and seam quality, needle lines
• Trims, including buttons, zips and their attachment strength
• Colour against the approved standard under specified lighting
• Labelling, including composition, care, origin and any required warnings, verified against the test results
• Packing, covering folding, polybags, hangtags, cartons and shipping marks

Measurement is where the highest-value findings usually appear. A grade error that leaves the largest size 4 cm short is invisible on a single approved sample but obvious across a measured set — which is why measurement checking against a full size set matters more than any single measurement on a single garment.

Specification Versus Verification

There is a structural point buried in the test matrix that most buyers miss. Almost every item in it is a specification decision that has to be made *before* production, not a verification step afterwards.

ISO 3758, the care labelling standard, is the clearest example. The care symbols on your label and the restrictions they imply — no tumble drying, cool iron, dry clean only — are determined by how the garment performs. If you specify the care label before the wash test, you have committed to a claim you have not verified, and a machine-washable label on a garment that felts is a false claim with a returns trail attached.

The same logic runs through the other tests. Pilling performance is set by fibre length and yarn structure. Colour fastness is set by the dyeing and the fibre. Fibre composition must be confirmed before labels are printed, because re-labelling a finished shipment is a rework project. Specify the performance you need, verify it with the test, then let the label follow the result.

FAQ

What does AQL 2.5 mean in a garment inspection?

AQL 2.5 is the acceptance quality limit most commonly applied to major defects in general consumer goods. It does not mean 2.5% of the order may be defective; it defines a statistically derived acceptance and rejection threshold for the sample size drawn from the lot. Critical defects are normally set at AQL 0.

How much shrinkage is acceptable for a knitwear order?

There is no universal figure — it depends on the fibre, construction and your own quality standard. What matters is that a limit is written into the PO, that it is tested using ISO 6330 for the wash procedure with ISO 5077 for measurement, and that the care label matches the tested behaviour.

Do I need both laboratory testing and a final inspection?

Yes. Laboratory testing verifies how the material performs — shrinkage, pilling, colour fastness, restricted substances. Final inspection verifies that the finished garments match the approved specification — measurements, workmanship, labelling, packing. They test different things and neither replaces the other.

Which pilling test should a sweater PO specify?

ISO 12945-2 (modified Martindale) is the common choice for knitwear, often paired with ISO 12945-4 for visual rating. On the US side ASTM D4970 is the Martindale-based equivalent. Specify the part number, not just the standard number, since ISO 12945 has four parts with different methods.

Sources

• ISO 6330:2021, Textiles — Domestic washing and drying procedures for textile testing
• ISO 5077:2007, Textiles — Determination of dimensional change in washing and drying
• ISO 12945-1 to -4:2020, Textiles — Determination of fabric propensity to surface pilling, fuzzing or matting
• ISO 12947-2:2016, Textiles — Determination of the abrasion resistance of fabrics by the Martindale method
• ISO 105-C06:2010, ISO 105-X12:2016 and ISO 105-B02:2014, Tests for colour fastness
• ISO 14184-1:2011, ISO 3071:2020, ISO 14362-1:2017 and ISO 1833 series
• ISO 2859-1:2026, Sampling procedures for inspection by attributes — AQL-indexed sampling schemes; ANSI/ASQ Z1.4
• AATCC test methods TM135, TM150, TM61, TM8, TM16.3, TM112 and TM81
• ASTM D4970 and D4966, pilling and abrasion test methods

The Bottom Line

Testing and inspection answer two different questions — how the material performs, and whether the garments match the spec — and a complete quality plan needs both. Specify the standard *and* the part, set critical defects at zero rather than blending them into a 2.5 calculation, decide the care claim after the wash result rather than before, and remember that the acceptance number at AQL 2.5 is more generous than intuition suggests. Quality is bought at the specification stage and only confirmed at inspection.

LINFAIR tests to your nominated standards through accredited laboratories, runs AQL inspections with photo and video reports at every production milestone, and specifies care labelling from wash-test results rather than assumption. Send us your quality standard with your tech pack and we will confirm the test plan before production starts.