The Garment Inspection Checklist: Pre-Production, Inline & Final

This is a detailed checklist. Not a lecture on why quality matters; you already know that.

It’s the list a factory works through at each of the three checkpoints above, rewritten for a founder who is placing an order and wants to know what “good” looks like at every stage.

Print it. Send it to your manufacturer. Use it on a video call.

The benefit is simple: a bulk order that arrives matching the sample you approved correct fabric, correct color, correct measurements, packed clean, with no nasty surprise when you open the first carton.

Here’s the idea that makes all three checkpoints worth the effort.

Checkpoint 1 · Before the run

Pre-production: check the inputs before anything is cut

This is the cheapest checkpoint and the one founders skip most.

Nothing has been sewn yet.

Every problem you find here is a problem that costs almost nothing to fix because you’re changing an input, not unmaking a finished garment.

Pre-production inspection has one job: confirm that what the factory is about to run with matches what you approved.

Three things get checked: the materials, the color, and the golden sample.

01 · Materials and fabric:

The garment can only be as good as the cloth it’s cut from.

Before bulk cutting starts, the fabric that arrived should be inspected against the spec you agreed on for composition, weight (GSM), width, hand-feel, and finish.

It should be the real bulk fabric, not a lookalike substitute.

Factories check incoming fabric with the 4-point system, the most widely used fabric-grading method in the industry. Standardized under ASTM D5430, it assigns 1 to 4 penalty points to each visible flaw based on its size, with no single flaw scoring more than 4 points.

Add up the points across the roll and convert them to points per 100 square yards, and you get a grade.

A common acceptance line is around 40 points per 100 square yards; above it, the roll is treated as “seconds.”

You don’t need to run this yourself. You need to know it exists and to ask for the fabric inspection report.

Before the run, confirm your fabric with the same care you’d give the finished garment.

Getting the cloth right is upstream of everything, and it starts long before this stage at sourcing and specifying the right fabric.

02 · Color and lab dips:

Color is the single most common thing to come back wrong across a run.

A lab dip approved on a screen, or under the wrong light, is not approved at all.

Before the run, the bulk color should be signed off against a physical reference, a Pantone chip, or your approved lab dip, in a controlled light booth, not by the window.

This deserves its own process, and locking color so it doesn’t drift is a topic in its own right.

The finished garments get verified against the same physical standard again at Checkpoint 3.

At this checkpoint, the box to tick is simple: is bulk color approved against a physical standard in correct lighting in writing?

Get that yes now, and the final color check becomes a formality rather than a crisis.

03 · The golden sample:

Everything at the next two checkpoints is measured against one object: the golden sample.

That’s the approved, signed, and sealed reference garment that defines exactly what bulk should be. It’s produced and locked during the sampling process.

Before the run begins, there should be a pre-production (PP) sample made from the real bulk fabric on the real production line with the real trims, and it should be approved and sealed.

If the factory wants to start bulk without a signed PP sample, that’s the moment to pause.

You’re about to lose your reference point.

Confirm the PP meeting has happened, and the sample is signed off and sealed as your production standard.

Checkpoint 2 · During the run

Inline: Catch defects while sewing is still happening

This is the checkpoint almost no first-time founder asks for, and it’s the one that saves whole runs.

Inline inspection, sometimes called DUPRO (during-production), happens while the line is still running, so a problem found now gets fixed on the remaining units instead of repeated across every one.

The logic sits right on the defect cost curve.

If a seam is puckering on the first 50 units, you want to know at unit 50, not at unit 2,000.

A run reworked mid-stream is contained.

A run reworked after it’s finished is a disaster.

The first-off-line check:

The first meaningful inline check is on the first garments to come off the line.

They’re compared against the golden sample: construction, seam types, stitch density, placement of prints and embroidery, symmetry, and trims.

If the first units are right, the run has a green light.

If they’re not, you’ve caught it at the earliest possible moment.

Where a run reworked mid-stream still stings is cost, and knowing what a production run actually costs to make is what tells you whether a rework is a shrug or a real problem.

Sampling, not 100%: how AQL works?

You can’t check every unit of a large run, and you don’t need to.

Factories and third-party inspectors use AQL, the Acceptance Quality Limit, to inspect a statistically valid random sample and make an accept/reject call for the whole lot.

The method is set out in ISO 2859-1 (its American twin is ANSI/ASQ Z1.4).

You pick an inspection level; General Inspection Level II is the default for most garment orders, and the standard’s tables convert your order size into a sample size and an accept/reject number.

Inspectors then sort every defect they find into three classes:

In plain terms, critical defects carry an AQL of 0; a single one fails the lot, while major and minor defects each get a ceiling.

Go over the ceiling, and the whole batch is rejected.

You don’t have to memorize the tables.

You need to agree on the AQL levels with your manufacturer before the run, so “pass” means the same thing to both of you.

The full mechanics, sample sizes, accept/reject math, and inspection levels are in our deeper guide to AQL sampling and acceptance limits.

The classes get less abstract with real garment examples.

A critical defect is a broken needle tip left in a seam or a drawcord that fails a child-safety pull test, a safety or legal problem, so zero are allowed.

A major defect is the kind that gets a garment returned: a hole, an open seam, a chest measurement 3 cm out of tolerance, or a print peeling on the first wash.

A minor defect is cosmetic and easy to miss: a stray thread, a slightly uneven topstitch, or a faint water spot that brushes off.

The same problem can even change class with severity: a tiny mark is minor; the same mark across the chest is major.

What does inline actually catch?

Inline is where process defects show up, the ones that repeat because something on the line is set up wrong.

Knowing the usual suspects helps you ask the right question.

Common recurring defects and how they’re graded are covered in our reference on common garment defects.

• Construction: skipped or broken stitches, seam pucker, open seams, wrong stitch density.
  

  ---

• Symmetry: pockets, panels, or prints that don’t line up left to right.
  

  ---

• Decoration: print cracking, misregistered screens, and embroidery pulling or puckering the fabric.
  

  ---

• Shade: panels cut from different dye lots ending up on the same garment.
  

  ---

• Trims: zippers that catch, drawcords the wrong length, and labels sewn in crooked.

Checkpoint 3 · After the run

Final: measure, count defects, and check packing before you pay the balance

Final inspection, also called pre-shipment inspection, happens when the order is finished and packed, typically at 80–100% complete.

It’s the last gate before the goods leave.

Three things get checked: measurements, defects, and packing.

01 · Measurements and tolerance:

Every garment style has a set of points of measure (POM): chest width, body length, sleeve length, and so on, each with an approved value and a tolerance (how far it can vary and still pass).

In the final, a sample of garments is measured across the full size run and checked against that spec.

Two things matter here. First, ensure individual garments sit within tolerance.

Second, that the fit holds across sizes and that the size grading is correct, so a large is genuinely a large.

The measurement points and grading logic behind this live in our guide to measurement specs and size grading.

02 · The defect audit:

This is where AQL does its final job.

The inspector pulls a random sample per the ISO 2859-1 tables, inspects each unit, sorts defects into critical/major/minor, and compares the totals to the accept/reject numbers you agreed on.

Over the ceiling on majors or minors or a single crit, and the lot fails. Under, and it passes.

Color is checked here too, one last time: the finished garments are compared against the same physical standard you approved at Checkpoint 1 to catch any drift that crept in during bulk dyeing.

If you locked it properly at the start, the full method is in our guide to lab dips, Pantone matching, and color consistency; this is a formality.

If you didn’t, this is where it surfaces, at the most expensive point on the curve.

03 · Packing and presentation:

The garments can be perfect, and the order can still fail here.

Final inspection checks that each piece is folded, tagged, and polybagged correctly; that hangtags and care labels are right; that the ratio of sizes and colors in each carton matches the packing list; and that cartons are labeled and sealed as agreed.

How your goods should be packed and cartoned is set out in our guide to packing, polybagging, and carton marking.

What happens if the lot fails?

A failed inspection isn’t a dead end; it’s the system working.

When a lot exceeds the agreed AQL, the standard path is that the factory reworks the defective units, and then the lot is re-inspected against the same standard.

Nothing ships until it passes.

This is exactly why the earlier checkpoints matter; a failure at the final one is recoverable, but it costs the most time, which is the whole point of the Defect Cost Curve.

Two things make a fail manageable rather than a fight.

First, the golden sample: with a signed physical reference, “this isn’t right” is a fact, not an opinion, so there’s nothing to argue about.

Second, your payment terms: if the balance is tied to a passed inspection, the factory is motivated to rework fast, and you’re not left chasing a refund on goods you’ve already paid for.

Both of those are decided long before the run, which is why the paperwork below is part of the checklist, not an afterthought.

The paper trail:

Documentation and photo evidence:

A checklist worked through in someone’s head disappears the moment there’s a dispute.

The thing that protects you isn’t the inspection; it’s the record of it.

At every checkpoint, the result should be written down and photographed.

The inspection report:

Each checkpoint should produce a short report:

What was checked was the sample size, the defects found by class, the measurements taken, and a clear pass or fail against the agreed AQL.

A dated, signed report turns “I think it looked fine” into a fact both sides accepted at the time.

Photo evidence:

Photos are the cheapest evidence there is.

At each checkpoint, capture the following: the golden sample (sealed and tagged), any defects found, the measurement check in progress, and the packed cartons before they ship.

If a shipment is disputed later, timestamped photos settle it faster than any email chain.

Why the record matters more than you think?

The documentation is what lets you tie your final payment to a passed inspection, the single most protective term you can put in a manufacturing deal.

Without a report, “passed inspection” is an opinion. With one, it’s a condition.

What your agreement should say about failed inspections and payment triggers is covered in our guide to getting your manufacturing terms in writing.

The tool:

The downloadable inspection checklist:

Here’s the whole thing in one place:

The three checkpoints plus documentation, in the order they happen.

Print it, or send it to your manufacturer before your next run so you’re both working to the same standard.

Questions founders ask