Short answer: Use the lot size and inspection level to find a code letter in the sample-size code-letter table. In the matching single-sampling table, find that letter's row and your AQL column. Read the sample size (n) and the acceptance and rejection numbers (Ac/Re), following any arrows. Inspect the sample, then compare the agreed count with Ac and Re for each defect class.
We are UTS, a third-party provider of product inspection, factory audit, supplier evaluation, and laboratory testing coordination. This guide explains a single-sampling table-reading method. Our reports present actual findings, not corrective-action plans, improvement recommendations, or shipment-release advice. The preparation guidance below is separate from report content.
What the AQL Table Tells You
AQL means Acceptance Quality Limit. ISO 2859-1 covers sampling schemes indexed by AQL for lot-by-lot inspection by attributes [1]. The AQL is an input to the sampling scheme, not permission for that percentage of every shipment to be defective.
Our final random inspection service describes AQL sampling under ANSI/ASQ Z1.4. Use the standard and edition specified in your inspection agreement; do not assume that similarly named tables from different standards or editions are interchangeable.
In a single-sampling plan, a random sample supplies the evidence for a lot-acceptance decision. NIST describes this as an (n, c) plan: inspect n items and reject the lot if more than c are defective [2]. Acceptance is a decision about the defined lot based on sample evidence, not proof that every unit meets requirements. Where every unit needs the specified checks, full inspection is a different scope from sampling.
Confirm the Inputs Before Reading
First confirm the governing standard, edition, and sampling type. The steps below use single sampling; double and multiple sampling require their own stages and decision rules. Then establish these inputs:
- Defined lot and quantity: use the goods covered by this inspection, not automatically the entire purchase order. Identify which models, sizes, or other variants belong to the lot and how the plan covers them.
- Inspection level: use the agreed general or special level. The level and lot size together locate the code letter; neither is an AQL value.
- AQL by defect class: identify the approved major and minor criteria, plus any separate critical-defect rule. Do not borrow another buyer's settings.
- Inspection severity: select the normal, tightened, or reduced table required by the applicable scheme and lot history. Do not change severity merely to obtain a preferred result.
- Counting basis: establish whether the plan counts nonconforming units or individual nonconformities. The table and the count must use the same basis.
UTS's final random inspection page lists Major 2.5, Minor 4.0, and zero tolerance for critical defects. These are service criteria to confirm for the order, not universal settings. “Critical 0” means no critical defects are accepted in the checked sample under that requirement; it is not an instruction to look for an AQL 0 column in the standard table.
Find the Sample-Size Code Letter
Open the sample-size code-letter table. Find the row containing your defined lot quantity, then move across to the agreed inspection-level column. Record the letter at their intersection.
The code letter links the lot-size table to the sampling-plan table. It is not a quality grade or an acceptance decision.
Check the lot-size range carefully at a boundary. A quantity outside the selected range means you are starting from the wrong row, even if every later lookup is correct.
Read n, Ac, and Re
Open the single-sampling master table for the agreed severity in the same standard and edition. Locate the code-letter row, read its sample-size entry, and move across to the AQL column for the defect class. The row supplies n; the AQL cell supplies Ac/Re or directs you elsewhere.
| Entry | Meaning |
|---|---|
| n | The sample size required by that plan. |
| Ac | The largest count accepted under the plan's acceptance rule. |
| Re | The smallest count meeting the plan's rejection rule. |
Repeat the AQL-column lookup for each defect class. Keep the resulting n, Ac, and Re together; do not assume every class necessarily uses the same sample size.
If a cell contains an arrow, follow it to the indicated plan and use that plan's sample size and decision numbers together. Read the table's legend and footnotes. Do not combine n from the original row with Ac/Re from the destination row or invent values for a blank cell.
Keep the Counting Basis Consistent
A nonconforming unit and a nonconformity are not the same count. Under a nonconforming-unit basis, one unit with two major defects counts as one major-nonconforming unit, not two. Under a nonconformity basis, the individual defects are counted according to that plan's definitions. These are different applications, not interchangeable ways to reach Ac.
The example below uses major-nonconforming units. Agree how units with findings in more than one defect class are recorded before applying separate class thresholds.
Apply the Acceptance and Rejection Rules
Select the required sample randomly from the defined lot, rather than inspecting only supplier-preselected units. NIST defines lot acceptance sampling around a random sample used to decide the disposition of the lot [3]. Complete the required checks before comparing the results with the plan:
- A count at or below Ac meets the acceptance rule.
- A count at or above Re meets the rejection rule.
- If the table leaves a count between Ac and Re or includes additional instructions, follow those instructions. Do not force it into a simple binary rule.
The following is a hypothetical illustration, not a value extracted from a standard: assume n is 200, Ac is 10, and Re is 11 for major-nonconforming units.
| Major-nonconforming units found | Outcome under this illustrative plan |
|---|---|
| 8 | Accept |
| 10 | Accept, at Ac |
| 11 | Reject, at Re |
| 14 | Reject |
Ten major defects spread across ten units and ten major defects found on one unit do not produce the same nonconforming-unit count. Replace all illustrative values with the approved plan, and evaluate every required defect class. Passing the major criterion alone does not establish that all inspection requirements were met.
Avoid These Table-Reading Errors
- Using AQL as the sample percentage: an AQL of 2.5 does not mean inspecting 2.5 percent of the lot. Read n from the sampling plan.
- Mixing inspection level and severity: the level helps locate the code letter; severity determines which sampling table applies.
- Switching tables without the rules: normal, tightened, and reduced inspection belong to a scheme with switching requirements. A blank booking field does not establish that any table is suitable.
- Extending the result beyond the defined lot: goods outside the sampling frame are not covered simply because they share a purchase order.
- Treating on-site sampling as laboratory evidence: laboratory work needs its own scope and criteria. ISO/IEC 17025 addresses the competence of testing and calibration laboratories; it is not an AQL sampling table [4].
How UTS Applies the Agreed Requirements
Our published final random inspection workflow starts with the buyer's purchase-order details, product specifications, and factory address. Our team prepares the inspection checklist with the product requirements, AQL levels, and special testing needs. The on-site work includes random sampling, physical testing, and visual inspection, with observed defects, measurement data, and photographs documented in the report.
Those materials connect the table to the product: Ac/Re supplies the count threshold, while the specification, defect definitions, and any approved sample reference establish what is being checked. A table alone cannot determine whether a finish, dimension, or function meets the buyer's requirement.
For table-reading preparation, send our team the standard and edition, lot quantity, inspection level, severity, AQL by class, counting basis, and defect definitions. Identify separately booked laboratory tests and any limits on access to the goods. Missing checks or inaccessible goods must not be presented as completed inspection coverage.
Where barcode or QR-code checks are within the agreed scope, the requirement is a 100% scan success rate for the checked barcodes or QR codes, with returned data matching the agreed reference using the specified equipment and conditions. This applies only to the checked items; it does not mean the entire lot was scanned or guarantee readability on every future device.
Our reports present actual inspection scope, completed checks, measurements, test results, observed defects, and checks that could not be completed. They do not contain sorting, rework, improvement, or shipment-release recommendations. Any later action remains the buyer's and supplier's decision.
Check Your Reading Before Inspection
Your completed lookup should identify the standard and edition, sampling type, lot size, inspection level, severity, code letter, and n/Ac/Re for each class, together with the counting basis and any applicable table notes. If one is missing, resolve it before selecting the sample rather than reconstructing the plan after seeing the results.
Preparing an AQL inspection? Share your lot details and approved inspection requirements with our team.
