AQL 2.5 for Activewear: Inspection Checklist and Defects

By Jerry · August 11, 2026 · Updated August 11, 2026

Short answer: AQL 2.5 is an indexing value used with an acceptance-sampling standard; it is not permission to ship 2.5% defective garments. A valid plan also needs a defined lot, current standard, inspection level, inspection status, sample size, acceptance and rejection numbers, random selection method, and written defect classifications. Inspect workmanship, measurements, function, labeling and packing against the approved specification, then record defects by severity before deciding whether to accept, hold, rework or reinspect the lot.

For Linked Sourcing, quality planning depends on the ordering route:

  • Ready Styles from 50 pieces per style: select an existing product and available size/color mix from the current private Excel catalogue. Ask Jerry for the stock quotation sheet, then agree what will be checked on the confirmed lot. Request the Ready Styles files.
  • Full Custom Manufacturing from 200 pieces per color and style: define the quality standard, defect list, measurement tolerances, test requirements and final-inspection plan before bulk production. Start a Full Custom quality review.

The following guide explains how to write that plan. It does not reproduce the copyrighted sampling tables in ISO or ANSI/ASQ standards, and it does not turn an indicative inspection convention into a contractual result.

What AQL 2.5 actually means

The current ISO 2859-1:2026 provides acceptance-sampling schemes for inspection by attributes, indexed by acceptance quality limit (AQL), for lot-by-lot inspection. ISO published this third edition in January 2026; it replaced the withdrawn 1999 edition and includes single, double and multiple sampling schemes plus switching rules.

The ASQ quality glossary describes AQL in a continuing series of lots as the limit of a satisfactory process average for sampling inspection. That is why AQL 2.5 is not the same as allowing 2.5% of one shipment to be defective.

A sample result produces a decision about a lot under a selected plan. It does not reveal the exact defect rate of every uninspected item. A good lot can occasionally be rejected and a poor lot can occasionally be accepted; the sampling plan manages those risks but does not remove them.

Never calculate an acceptance number by multiplying the order quantity by 2.5%. The acceptance and rejection numbers come from the selected plan after all required inputs are known.

Why “inspect at AQL 2.5” is incomplete

An inspector cannot execute a defensible final inspection from that sentence alone.

Required inputQuestion the buyer must answerWhy it matters
Standard and editionISO 2859-1:2026, ANSI/ASQ Z1.4, another agreed plan, or a buyer manual?Different documents and editions must not be mixed
Lot definitionWhich finished units, colors, sizes, production lines and dates form one lot?Lot size drives the plan and prevents selective presentation
Inspection situationContinuing series or isolated/short-series lot?The appropriate sampling system can differ
Inspection levelWhich general or special level applies?Controls the relative amount of inspection
Inspection statusNormal, tightened, reduced or another agreed state?Switching rules change the plan over time
Plan typeSingle, double or multiple sampling?Changes how many samples and decisions can occur
Counting methodNonconforming units or nonconformities per 100 units?One garment can contain more than one defect
Defect classes and AQLsWhich defects are critical, major or minor, and what value applies to each class?A single “2.5” does not classify every problem
Random selectionHow are cartons and garments chosen across the lot?Factory-selected “best pieces” invalidate the sample logic
DispositionWhat happens after rejection or an observed critical issue?The standard gives a sampling decision, not the full commercial remedy

ISO 28590:2017 introduces the ISO 2859 family and guides selection of the appropriate attribute-sampling system. For an isolated lot or a series too short for the switching rules in ISO 2859-1, ISO 2859-2:2020 provides a separate system indexed by limiting quality (LQ). The buyer, quality professional or appointed inspection company should select the system that matches the real sourcing situation.

AQL, inspection level and defect class are separate decisions

Three labels are often collapsed into one:

  1. Inspection level helps determine how much of the lot is sampled.
  2. AQL value indexes the acceptance-sampling plan for a defined class or count.
  3. Defect classification determines which count a finding enters.

Changing one does not automatically change the other two. “General inspection level II” does not mean “AQL 2.5,” and “AQL 2.5” does not define whether an open seam is major or critical.

If a buyer writes separate values for critical, major and minor defects, treat them as three different acceptance criteria. ISO does not automatically assign one universal severity table to activewear. The approved product specification must say how findings are classified.

Critical, major and minor activewear defects

Use a product-specific defect library with photographs and written boundaries. The examples below are starting questions, not automatic classifications; intended use, destination law, buyer requirements and severity can change the result.

ClassPractical decision testActivewear examples to define before inspection
CriticalCould create an unacceptable safety, legal or serious-use risk?Broken needle or sharp point; prohibited component; dangerous cord/trim; serious mandatory-label error; contamination; a failure that makes the product unsafe for its stated use
MajorIs it likely to impair function, saleability or the approved appearance?Open or weak stretch seam; hole; severe stain; wrong size/color/style; measurement outside a critical tolerance; broken zipper; missing component; incorrect logo; agreed opacity or support failure
MinorIs it a limited departure that does not materially impair use or saleability?Small workmanship variation, limited stitch waviness, minor thread trimming or slight appearance issue within the buyer’s defined boundary

Avoid vague rules such as “all loose threads are minor.” A thread end can be cosmetic in one place and evidence of an unsecured seam at a stress point in another. Classify the failure mode, not only its visible name.

Also distinguish:

  • a nonconforming item: a garment with one or more specified nonconformities;
  • a nonconformity: one failure to meet a requirement.

If one legging has a stain and a broken seam, the chosen plan must say whether the decision uses defective garments or counts of individual nonconformities. Do not change counting methods after seeing the result.

The 12-part activewear final inspection checklist

1. Inspection identity and lot scope

Record the supplier, factory, purchase order, style, lot quantity, colors, sizes, carton count, production status, inspection date and location. State whether goods are finished and packed, and identify any excluded or unfinished units.

Do not combine different styles or materially different production runs only to reach a convenient lot size unless the agreed plan permits it.

2. Approved reference documents

The inspector should receive one controlled package:

  • purchase order and final order breakdown;
  • latest approved tech pack and bill of materials;
  • approved pre-production sample or sealed reference;
  • finished-garment measurement chart and POM diagram;
  • measurement tolerances;
  • approved colors, lab dips, prints and logo artwork;
  • label, barcode, packaging and carton specifications;
  • product test requirements and available reports;
  • defect classification library;
  • agreed sampling standard, edition and plan inputs.

Use the activewear tech pack checklist and size grading and fit approval checklist before bulk production so inspection does not become the first time conflicting instructions are discovered.

3. Random sample selection

Select cartons and garments across the lot, including different pallet/carton positions, colors and sizes as required by the plan. The sample should not be preselected by production staff.

Record which cartons were opened and how units were chosen. If the available goods do not represent the declared lot, pause and report the limitation instead of producing a confident pass decision from a biased sample.

4. Quantity, assortment and identification

Check ordered versus packed quantity, color/size ratio, SKU, barcode, hangtag, polybag or alternative packaging identification, carton markings and destination information. Scan representative barcodes where required rather than only reading the printed numbers.

5. Workmanship and seams

Inspect inside and outside under consistent lighting. Check:

  • skipped, broken, loose or unsecured stitches;
  • open seams, seam grin and raw-edge exposure;
  • stitch density and seam type against the approved construction;
  • puckering, twisting and left/right asymmetry;
  • bartacks and reinforcement at stress points;
  • flatlock, coverstitch, bonding or seam-tape continuity;
  • needle damage, cuts, holes and fabric caught in seams;
  • thread color, trimming and cleanliness.

For stretch seams, visual appearance alone is not enough. Apply the agreed extension or function check and record any popping, opening or failure.

The activewear seam strength and stretch test checklist explains why broken thread, seam grin, fabric rupture, woven-yarn slippage and stitch unravelling need different evidence and corrections. Final inspection can find construction drift, but an inspector’s hand pull is not a named seam-force or fabric-burst method.

6. Measurements and size identity

Measure the agreed POMs using the same conditioning, placement and tape method defined in the tech pack. Record actual values, not only “pass.” Verify that the physical size label matches the measured/specification size.

Measurement inspection may use a different sample quantity or rule from visual workmanship inspection. State that plan separately. AQL cannot repair unclear POM methods or unrealistic tolerances.

7. Fabric, shade and surface

Check fabric reference, hand feel, shade consistency, panel-to-panel color, print placement, stripes or repeats, contamination, odor, stains, snags, holes, barre, abrasion and other agreed surface requirements.

Visual inspection cannot verify fiber content, restricted substances, recycled content or every performance claim. Use current documents and laboratory tests where those requirements apply.

8. Stretch, recovery, opacity and movement function

Perform only checks that have a written method and pass condition. Depending on the product, these can include:

  • waistband or underband extension and recovery;
  • seam extension;
  • strap security and adjustment;
  • pocket opening and retention;
  • zipper, snap, drawcord and closure function;
  • pad placement and removal;
  • garment symmetry after extension;
  • opacity at an agreed stretch or fit condition;
  • basic movement checks against the approved reference.

GSM or elastane percentage alone cannot prove these outcomes. Use the activewear fabric weight and stretch guide to define material and sample evidence before final inspection. The activewear opacity and squat-proof testing guide explains how a controlled garment check can be repeated on sampled bulk units without misrepresenting AQL 2.5 as a percentage of “squat-proof” garments.

9. Logos, prints and trims

Check artwork, size, color, orientation and measurable placement. Inspect adhesion, cracking, peeling, embroidery backing, silicone parts, reflective transfers, zippers, elastics, cords, hardware and other trims against the approved sample.

A simple rub or stretch check during inspection is not a substitute for an agreed wash, abrasion or durability test.

Use the activewear pilling and abrasion test checklist to keep pilling, fuzzing, snagging and abrasion as separate failure questions and to record the full test conditions. Final visual inspection can compare bulk to an approved reference; it cannot convert a laboratory rub count into garment lifespan.

10. Labels and market requirements

Verify the exact approved content, language, location, attachment, orientation and legibility of main, size, fiber-content, care, country-of-origin and warning labels as applicable. Check that the physical garment, hangtag, barcode and carton identify the same SKU.

Legal requirements vary by destination and product. The factory or inspector should not invent mandatory wording; the buyer should supply approved market information or obtain qualified compliance advice.

11. Needle, sharp-point and contamination controls

Where required, review broken-needle records, tool control, metal-detection status and corrective records. Visually inspect for sharp points, fragments, oil, mold, insects, moisture damage and other contamination.

A safety-critical concern should trigger the agreed escalation even if the ordinary major/minor count would otherwise pass.

12. Packaging, cartons and report evidence

Check folding, individual protection, labels, assortment, carton dimensions/weight limits, sealing, shipping marks and product protection. Photograph carton condition, opened samples, defect examples and critical approvals.

The final report should state:

  • lot and sample identification;
  • standard edition and sampling inputs;
  • sample size and acceptance/rejection criteria by class;
  • defect count, classification and affected units;
  • measurements and functional results;
  • photos tied to defect references;
  • quantity and packaging findings;
  • pass, fail, pending or hold decision;
  • unresolved limitations and required follow-up.

AQL inspection does not replace these controls

Final random inspection is one layer, not the whole quality system.

ControlWhat it catches earlier or differently
Incoming material control (IQC)Wrong or unstable fabric, trims, shade and material documentation before cutting
In-process quality control (IPQC)Sewing, bonding, placement and construction drift while it can still be corrected
Laboratory/performance testingChemical, colorfastness, dimensional, strength and claim-related results that visual inspection cannot prove
Pre-production approvalThe final specification, fabric, trim, fit, branding and packaging reference
Final acceptance sampling (FQC)Lot-level shipment decision on a random finished sample under the agreed plan
100% check for selected featuresCritical or easily checked requirements where sampling risk is not acceptable

Calling all three stages “AQL 2.5 inspection” hides their different purposes. AQL-indexed acceptance sampling is most directly relevant to a defined lot and sampling decision; IQC and IPQC should use their own material and process controls.

For incoming fabric rolls, use the activewear fabric inspection and four-point system checklist to define the visual grading option, roll coverage, defect map, usable width, GSM, shade, bow/skew, quarantine and cut-release record. A roll point result cannot be converted into a finished-garment AQL result.

What to do when the lot fails

Do not ask the inspector to quietly replace defective sample pieces and continue. Preserve the original result and place the affected lot on hold.

A written disposition can include:

  1. reject the lot;
  2. sort 100% for specified defects;
  3. rework or replace affected units;
  4. investigate root cause and issue corrective action;
  5. present a clearly defined corrected lot for reinspection;
  6. accept a documented buyer concession for a specific issue.

The reinspection plan, cost responsibility and whether reworked units form a new lot should be agreed before the second visit. A concession on one order does not automatically change future tolerances or classifications.

A grip-sock inspection plan should classify missing high-contact deposits, under-cure/tack, sole-print registration, toe-seam opening, size-pair mismatch, grip transfer and incorrect safety wording by actual user/claim risk. Use the grip sock manufacturing guide to separate bulk workmanship checks from surface traction, fit, wash durability and chemical evidence.

A heavyweight-hoodie plan should classify wrong body/rib material or GSM, unapproved wash variation, shade mismatch, hood or seam twist, zipper failure/wave, rib distortion, pocket-end damage, decoration defects, drawcord hazards and incorrect labels by actual risk. Use the heavyweight hoodie manufacturing guide to connect final workmanship sampling to earlier material, wet-process, care and product-safety controls.

A heavyweight-jogger plan should classify wrong body/rib/elastic/drawcord material, unapproved wash or set shade, waist failure, rise/leg measurement error, seam twist, pocket/cuff/hem damage, decoration defects and incorrect labels by actual risk. Use the heavyweight jogger manufacturing guide to keep final workmanship sampling separate from earlier knee-recovery, fit, load, wash and claim validation.

Ready Styles inspection brief

Ready Styles use an existing product, standard fabric and available size range. Start with the actual confirmed stock or production lot rather than a theoretical public price list.

Hi Jerry, please send the current Ready Styles Excel catalogue and stock quotation sheet. Product type: ____. Estimated quantity: ____ pieces per style. Target market: ____. After I select styles, please confirm the available colors/sizes, lot quantity, current inspection records and options for final inspection or third-party inspection.

Ready Styles start from 50 pieces per style, with at least 15 pieces for each selected color and size combination. For a small isolated lot, agree whether a standard AQL scheme, a tailored sample, or 100% checking of selected features gives the right protection. Request the current Ready Styles files.

Full Custom quality brief

For Full Custom, approve the defect rules before the purchase order moves into bulk.

Hi Jerry, I want to develop a custom ____. Quantity: ____ pieces per color and style. Target market: ____. My required inspection standard/edition is ____. Proposed inspection level and plan: ____. Critical/major/minor criteria: ____. Measurement and test requirements: ____. Third-party inspection requirement: ____. Please review the quality plan before bulk production.

Full Custom Manufacturing starts from 200 pieces per color and style. Special materials, trims, constructions or testing programs can require a higher minimum or additional cost. Start a Full Custom inquiry.

Standards and source scope

This guide uses current official sources for the statistical framework:

  • ISO 2859-1:2026 — AQL-indexed acceptance-sampling schemes for lot-by-lot inspection; replaces the withdrawn 1999 edition;
  • ISO 28590:2017 — introduction to the ISO 2859 family and system selection;
  • ISO 2859-2:2020 — LQ-indexed plans for isolated lots or series too short for ISO 2859-1 switching rules;
  • ANSI/ASQ Z1.4 information and the ASQ quality glossary — current ASQ descriptions of attribute acceptance sampling, AQL, acceptance numbers and rejection numbers.

These standards do not supply Linked Sourcing’s commercial remedy, one universal activewear defect library or permission to mislabel a safety issue as minor. The purchase order and approved quality manual must define the product-specific requirements.

FAQ

What does AQL 2.5 mean in garment inspection?

AQL 2.5 is an indexing value used with an acceptance-sampling scheme. It helps select the applicable sampling plan and acceptance/rejection criteria after the standard edition, lot size, inspection level, inspection status and defect class are defined. It is not a complete inspection instruction by itself.

Does AQL 2.5 mean a factory can ship 2.5% defective garments?

No. Do not multiply the lot quantity by 2.5% to create an allowed defect count. AQL describes the limit of a satisfactory process average for a continuing series of lots in the sampling context. The pass/fail decision comes from the chosen standard's sample size and acceptance/rejection numbers, and any sampling decision carries producer and consumer risk.

What is the difference between a critical, major and minor defect?

The buyer's approved defect list should define the difference for the actual product and market. A critical defect can create an unacceptable safety, legal or serious-use risk; a major defect is likely to affect use, saleability or the intended appearance; a minor defect is a smaller departure that does not materially impair use. These are decision rules, not labels the inspector should invent after finding a problem.

What sample size should I use for a 500-piece activewear order?

The lot quantity alone cannot determine a valid sample size. You must also choose the current sampling standard, inspection level, normal/tightened/reduced status, single/double/multiple plan and AQL for each defect class. Use the licensed current standard or an inspection company working to the agreed edition; do not copy an unexplained online AQL chart.

Can a 50-piece Ready Style order use AQL inspection?

Yes, but the buyer should check whether a continuing-lot AQL scheme is appropriate for that small or isolated order. The sampling effort may be large relative to the lot, and critical or easily checked requirements can justify 100% verification. Ready Styles start from 50 pieces per style; agree the inspection scope after Jerry confirms the actual colors, sizes and available quantity.

What happens when an AQL inspection fails?

A failed result normally places the lot on hold; it does not automatically prescribe one commercial remedy. The buyer and supplier should agree whether to reject, sort 100%, rework, replace affected units, investigate root cause and reinspect a clearly defined lot. Any buyer concession should be written and should not silently change the defect standard for future orders.

Jerry — Linked Sourcing supply-chain specialist

Jerry · Supply-chain specialist, Linked Sourcing

Your English-speaking contact to a 40-year, 6-factory activewear group in China — from first sample to shipment. Connect on LinkedIn.

Two clear sourcing routes

Ready to source your next collection?

Request the current ready-style Excel catalogue from 50 pcs per style, or discuss full custom manufacturing from 200 pcs per color and style.