AQL Sampling Explained: AQL 1.0, 2.5, and 4.0

AQL stands for Acceptable Quality Limit, and it is the number that decides whether your shipment passes or fails inspection. When you set AQL 2.5 for major defects, you are telling the inspector exactly how many defective units in the sample are acceptable before the whole lot gets rejected. The three numbers importers use most are 1.0, 2.5, and 4.0, and they map to defect severity: 1.0 for the strictest categories, 2.5 for major defects, 4.0 for minor cosmetic issues.

If you have ever received an inspection report and wondered where the "pass" verdict came from, this is the mechanism behind it. AQL sampling levels translate your quality expectations into a concrete rule your inspector applies on the factory floor. CN Ally runs pre-shipment inspections against the AQL criteria written into purchase orders, which is where the right levels prevent disputes and the wrong ones create them.

What AQL actually measures

Here is the definition most importers get wrong: AQL is the worst tolerable quality level for a process average, expressed as a percentage of defective units. It is not the percentage of your lot that gets inspected, and it is not a promise that every unit shipped is defect-free.

Two things make this worth understanding. First, AQL works on samples, not on the entire lot. If you order 10,000 phone cases, nobody inspects all 10,000. The inspector pulls a statistically determined sample, counts the defects in it, and judges the whole lot from that sample. Second, the AQL number is the acceptance criterion applied to that sample, taken from the ISO 2859-1 sampling tables, which trace back to the MIL-STD-105E military standard and are still maintained through ASTM E2234.

The practical consequence: AQL is a risk-management tool, not a guarantee. A lot that passes at AQL 2.5 can still contain defective units. What the standard gives you is a defined, defensible line between "ship it" and "fix it," one that both you and the factory can agree on before production starts. That agreement is the whole point. Without written AQL criteria, every failed inspection turns into an argument about what "acceptable" means.

The three defect classes every AQL plan needs

Before the numbers mean anything, you have to sort defects into classes. An inspection checklist that treats a cracked housing and a loose thread as equal problems is useless. The industry standard sorts defects into three classes, and each class gets its own AQL value:

Defect class · What it covers · Typical AQL

• Critical: Defects that can harm the user or violate regulations: exposed wiring, sharp edges, toxic materials, missing safety certifications · 0
• Major: Defects the end customer would consider unacceptable: wrong function, broken parts, major finish flaws, incorrect dimensions · 2.5
• Minor: Departures from specification most buyers would tolerate: small scratches, slight color variation, loose threads · 4.0

Critical defects get an AQL of zero because there is no acceptable number of them. One exposed wire in a batch of electric kettles is one too many. Major defects get 2.5, which has become the default across apparel, home goods, toys, and general consumer electronics. Minor defects get 4.0, the loosest commonly used value, and it applies strictly to cosmetic issues.

This three-class structure is the combination most inspection firms use as their default: 0 / 2.5 / 4.0. Jonble, a China inspection provider, documents this as its standard setup and notes that buyers of high-risk products such as automotive components tighten it to something like 0 / 1.5 / 2.5. The structure stays the same; only the numbers move.

What AQL 1.0, 2.5, and 4.0 mean in practice

Each AQL value answers a different question: how much imperfection will you tolerate in this category before rejecting the lot? The numbers look abstract until you see the accept and reject counts they produce. For a 200-unit sample at General Inspection Level II, the standard tables give these results:

AQL value · Defects allowed in a 200-unit sample · Lot fails at

• 1.0: 5 or fewer · 6 or more
• 2.5: 10 or fewer · 11 or more
• 4.0: 14 or fewer · 15 or more

Notice that these are not percentages of the sample. AQL 2.5 does not mean 2.5 percent of the sample may be defective. It is a statistical parameter that produces these accept numbers through the sampling tables. This is the single most common misunderstanding we encounter.

So when should you use each level?

AQL 1.0 is the strict option. It suits defect categories where failure is expensive: products with safety implications, regulated goods, luxury items where the brand cannot absorb visible flaws, or importers selling into markets with aggressive return policies. The tradeoff is real: tighter AQL values drive more lot rejections and higher inspection costs, and suppliers often push back. As TradeAider noted in its 2026 comparison of the three levels, most reputable suppliers accept 2.5 without pricing adjustments, but dropping to 1.0 frequently prompts renegotiation.

AQL 2.5 is the workhorse. It is the default for major defects across most consumer product categories, and it is what your inspection firm will assume if you do not specify otherwise. For a first-time importer with no strong reason to deviate, 2.5 for majors is the sane starting point.

AQL 4.0 is for minor cosmetic defects only. Small scratches, slight color variations, minor finish imperfections. The one rule that matters here: never apply 4.0 to functional defects. Permitting that many functional failures through sampling is how importers end up with containers of products that look fine in photos and do not work in customers' hands.

How sample sizes are determined from your lot size

The AQL number is only half the equation. The other half is the sample size: how many units the inspector actually pulls from your lot. That number is not chosen by feel. It comes from a two-step lookup in the sampling tables.

Step one: find your sample size code letter. The tables cross-reference your lot size against the inspection level to produce a letter. The standard choice is General Inspection Level II, which most inspection firms use by default. Level I is less stringent (smaller samples, cheaper inspections, more risk), and Level III is the most stringent (larger samples for high-stakes products). Here is how lot sizes map to sample sizes at Level II:

Lot size (units) · Sample size at Level II

• 2 – 8: 2
• 9 – 15: 3
• 16 – 25: 5
• 26 – 50: 8
• 51 – 90: 13
• 91 – 150: 20
• 151 – 280: 32
• 281 – 500: 50
• 501 – 1,200: 80
• 1,201 – 3,200: 125
• 3,201 – 10,000: 200
• 10,001 – 35,000: 315
• 35,001 – 150,000: 500
• 150,001 – 500,000: 800
• 500,001 and over: 1,250

Step two: cross-reference the sample size against your chosen AQL value to get the accept and reject numbers. A 200-unit sample at AQL 2.5 gives an accept number of 10. A 315-unit sample at AQL 2.5 gives 14. The tables do the statistics; you just need to know which inputs produce which outputs.

One detail worth knowing: sample size grows much more slowly than lot size. A lot of 3,500 units gets a 200-unit sample, and a lot of 35,000 gets 315. Beyond a certain point, inspecting more units adds cost without meaningfully improving the decision.

A worked example: 3,500 units at 0 / 2.5 / 4.0

Let us walk through a realistic order. You have ordered 3,500 Bluetooth speakers. Production is finished and 100 percent of the goods are packed. You book a pre-shipment inspection and set the standard criteria: 0 for critical defects, 2.5 for major, 4.0 for minor.

First, the sample size. Your lot of 3,500 units falls in the 3,201–10,000 band, so at General Inspection Level II the inspector pulls 200 units at random from across the packed cartons. Random matters: the inspector must not let the factory hand-pick the sample.

Next, the accept numbers for a 200-unit sample, straight from the ISO 2859-1 tables:

• Critical defects at AQL 0: the lot fails if even one critical defect appears. A speaker with exposed wiring or a swollen battery ends the inspection on the spot.
• Major defects at AQL 2.5: the lot passes with 10 or fewer defective units in the sample, and fails at 11. A dead-on-arrival speaker, a non-functional button, or a badly misaligned housing counts here.
• Minor defects at AQL 4.0: the lot passes with 14 or fewer defective units, and fails at 15. Light scratches on the casing or a slightly off-center logo land in this bucket.

Suppose the inspector finds 0 critical defects, 8 major defects, and 12 minor defects in the 200-unit sample. The verdict: pass, on all three classes. But if the count is 0 critical, 13 major, and 9 minor, the verdict is fail, because the major-defect count exceeds the accept number of 10. One class failing fails the lot.

What happens after a fail depends on what your contract says, which is why the AQL criteria belong in the purchase order before production begins, not in an email after the inspection report lands.

Mistakes importers make when setting AQL

Most AQL problems are not statistical. They are judgment errors made when writing the purchase order. These are the ones we see most often.

Setting everything to 1.0 out of caution. It feels safe, but an AQL that is too strict produces failed inspections, re-inspection fees, delayed shipments, and a supplier relationship that turns adversarial. Strictness has a price, and the factory will eventually pass it back to you in higher quotes.

Setting everything to 4.0 to keep the peace. The mirror image. A loose AQL on major defects means lots full of functional problems sail through inspection with a passing report. If your return rate climbs after you "fixed" quality with inspections, check the AQL values in your purchase orders before blaming the inspector.

Applying 4.0 to functional defects. The 4.0 level exists for cosmetic issues. The moment a defect affects whether the product works, it belongs in the major class at 2.5 or tighter. Sorting defects into the right class matters more than fine-tuning the numbers.

Confusing the inspection level with the AQL value. These are separate dials. The inspection level (I, II, III) controls how many units get sampled. The AQL value controls how many defects are tolerated. An importer who demands "Level III inspection" thinking it means stricter acceptance has changed the sample size, not the pass criteria. Set both deliberately.

Leaving defect classification to the inspector. If your purchase order says "AQL 2.5" with no defect list, the inspector decides on the spot which defects are major and which are minor, and different inspectors draw those lines differently. A product specification with example defects in each class, agreed with the supplier before production, removes the ambiguity. Our complete QC guide covers how inspection fits into the wider quality program.

How to state AQL in your purchase contract

AQL criteria only protect you if they are written into the commercial documents before money changes hands. A verbal understanding with the sales rep does not survive a failed inspection. The clause needs four elements:

1. The defect classification. List example defects for each class, or attach your full product specification with defects marked. "Critical: any safety-related defect, wrong voltage, missing certification marks. Major: any functional failure, wrong dimensions beyond tolerance, incorrect labeling. Minor: cosmetic imperfections not affecting function."
2. The AQL values per class. State them explicitly: "Critical defects: AQL 0. Major defects: AQL 2.5. Minor defects: AQL 4.0, per ISO 2859-1 single sampling, General Inspection Level II."
3. The inspection standard. Name it: "Inspection per ANSI/ASQ Z1.4 (ISO 2859-1), single sampling plan, normal inspection." This tells the inspection firm exactly which tables to use and stops arguments about methodology.
4. The consequence of failure. "Final payment is released only after the goods pass pre-shipment inspection against the above criteria. Failed lots will be reworked or reproduced at the supplier's expense, followed by re-inspection." This is the clause that gives the AQL numbers their teeth. Tying it to your payment terms is what turns a quality preference into an enforceable commercial term.

Keep the wording plain. You are making sure a factory manager, an inspector, and you all read the same rule the same way. If a supplier refuses to sign AQL criteria into the purchase order, treat that as information about how they expect the inspection to go.

Frequently asked questions

What does AQL 2.5 mean in simple terms?

It means that for the defect class you applied it to, the sampling tables allow a defined number of defective units in the inspection sample before the lot fails. For a 200-unit sample, that number is 10. It does not mean 2.5 percent of your shipment may be defective. It is a statistical acceptance parameter, not a defect quota.

How many units will the inspector check from my order?

It depends on your lot size and inspection level. At the standard General Inspection Level II, a lot of 1,000 units gets an 80-unit sample, a lot of 3,500 gets 200 units, and a lot of 12,000 gets 315 units. Your inspection firm calculates this from the ISO 2859-1 tables when you book.

Can I just set AQL 1.0 for everything to be safe?

You can, but expect more failed lots, re-inspection costs, shipment delays, and suppliers who raise prices to cover the rework. Strict AQL values suit safety-critical and regulated products. For ordinary consumer goods, they usually cost more than the defects they prevent.

What happens if the defect count exceeds the AQL limit?

The lot fails inspection. What follows depends on your purchase contract: typically the supplier reworks or reproduces the defective goods at their expense, followed by re-inspection. This is why the AQL criteria and the failure consequences belong in the purchase order before production, not negotiated after a failed report.

Is AQL 0 the same as demanding zero defects?

No. AQL 0 for critical defects means the lot fails if any critical defect appears in the sample. It does not mean the entire shipment is defect-free, because only a sample was inspected. It is the strictest practical setting, and it is the right one for defects that can injure users or break regulations.

Should I use a different inspection level instead of the standard Level II?

Level II is the default for good reason: it balances inspection cost against decision quality for most products. Consider Level III for high-value or high-risk shipments where a larger sample is worth the extra inspector time, and Level I only for low-risk, low-value goods. Change the level deliberately, not by accident.

Setting your AQL: a decision rule

Start with the industry default of 0 for critical defects, 2.5 for major, and 4.0 for minor, at General Inspection Level II. Deviate only for a stated reason: tighten to 1.0 or 1.5 where defects carry safety, regulatory, or brand risk; loosen a class only where you have genuinely decided the defect does not matter to your customer. Write the defect classes, the values, the standard, and the failure consequences into your purchase order before production starts. Then let the tables do their job.

If you would like help setting AQL criteria for a specific product or writing them into your purchase orders, contact us at hi@cnally.com. Getting the numbers right before production is the cheapest quality investment in the whole sourcing process.