AQL sample size calculator for garment inspection

It looks the answer up in two tables from ISO 2859-1; it does not calculate anything. The first table turns your lot size and inspection level into a sample size code letter: 5,000 pieces at General Level II is letter L. The second table, the single sampling plans for normal inspection, gives each letter a sample size and, under each AQL, an accept and a reject number: letter L means 200 pieces, with 10 accepted and 11 rejected at AQL 2.5, and 14 and 15 at AQL 4.0. The sample depends on the lot size band, not the exact quantity, so 3,201 pieces and 10,000 pieces both get a sample of 200. We typed the tables in from the printed standard, and every time the site is built a check compares them with the figures in our AQL guide and stops the build if one differs. Always confirm against the table your inspector or your retailer’s manual uses.

Because some cells in the printed table hold an arrow instead of numbers, where the table has no plan for that combination of code letter and AQL. The instruction printed with the table is to use the first sampling plan the arrow points to, with that plan’s sample size as well as its accept and reject numbers. A lot of 100 pieces at General Level II is letter F, a sample of 20. At AQL 2.5 the table gives numbers for F: accept 1, reject 2. At AQL 1.5 it has a down arrow, so the plan becomes letter G: 32 pieces, accept 1, reject 2. On a small lot, major and minor defects can therefore end up with different sample sizes, and the calculator shows each one and says which arrow it followed. If a plan’s sample is as large as the lot, the instruction is to inspect every piece.

Enter the pieces of one style that will be offered for inspection together. The standard applies to lots from one producer using one process, so the same style made in two factories is two lots, each with its own sample. Different garments, such as a hoodie and a t-shirt on one purchase order, are normally inspected as separate lots too. The colours and sizes of one style are usually one lot, with the sample drawn at random across colours, sizes and cartons. How you split matters: one lot of 5,000 pieces needs a sample of 200 at General Level II, while two lots of 2,500 need 125 each, 250 in total. If a style ships in two parts, each part offered for inspection is usually treated as its own lot. Write the lot definition, the inspection level and the AQLs into the purchase order before production, so the factory, the inspector and you count the same thing.

On most lots it changes only the accept number, because the sample size comes from the lot size and the inspection level, not from the AQL. In our AQL calculator, 5,000 pieces at General Level II is always code letter L and 200 garments: the lot passes with up to 10 defects at AQL 2.5, 7 at 1.5, 5 at 1.0 and 3 at 0.65, and with up to 14 at 4.0. A 1,000-piece lot is 80 garments, with 5, 3, 2 and 1 accepted. The exception is a lot where the ISO 2859-1 table prints an arrow instead of numbers, which happens mostly on small lots: 500 pieces at AQL 0.65 moves from 50 garments to 80, accept 1. If you want more garments checked, change the level, not the AQL: Level III raises the 5,000-piece sample to 315, with 14 accepted at AQL 2.5. Choose both before production and write them into the purchase order.

Yes for normal single sampling, because the basic tables are the same, but confirm against the copy named in your contract. The e-Handbook of the US National Institute of Standards and Technology, read in October 2026, records that MIL-STD-105D was adopted as ANSI Z1.4 in 1971 and, with minor changes, as ISO 2859 in 1974, and that the basic tables remain the same. The Bureau of Indian Standards lists IS 2500 (Part 1):2000 as active in October 2026 and identical, under dual numbering, to ISO 2859-1:1999. ISO replaced that edition with ISO 2859-1:2026 in January 2026, and the main changes listed in its foreword concern skip-lot inspection and operating characteristic curves, not the code letters or acceptance numbers. So the AQL calculator’s answer for 5,000 pieces at General Level II and AQL 2.5, which is 200 inspected, accept 10 and reject 11, should match each of them. Write the standard and its year into the purchase order.

Neither: the AQL calculator gives the accept and reject numbers, which the standard applies to either count, so settle it before the inspection. The introduction to ISO 2859-1:2026 says the standard applies either by counting the nonconforming items or by counting the nonconformities, and MIL-STD-105E of 1989, the public US standard with the same basic tables, lets AQLs of 10 or less be expressed either way. When garments are counted, that standard defines a major defective as a unit with one or more major defects, which may also have minor ones but no critical defect, and a minor defective as a unit with minor defects but no major or critical defect. So a shirt with one major fault and two minor ones counts once, as a major. Counting every fault separately is stricter against the same accept numbers. State the method in your purchase order next to the AQLs, and check that the inspection report follows it.

In proportion to how many pieces of each colour and size are in the lot, with the garments inside each group picked at random. That is the rule for representative sampling in clause 4.5.1 of MIL-STD-105E, the 1989 public US standard that shares its basic tables with ISO 2859-1: where appropriate, the number of units taken from each part of a lot matches that part’s share of it, and the units within each part are selected at random. The AQL calculator gives one figure for the whole lot, so the split is a second step. For 5,000 t-shirts, 3,000 black and 2,000 white, a sample of 200 means 120 black and 80 white. A size ratio of 1:2:3:2:1 then puts about 13, 27, 40, 27 and 13 black garments in the sample from S to XXL, drawn from cartons across the packing list. Ask the inspector to record the split by colour and size in the report.

Enter the unit your inspector will judge as one item, and use that unit for the lot size and the defect count. MIL-STD-105E of 1989, the public US standard that shares its basic tables with ISO 2859-1, defines the unit of product as the thing inspected, which may be a single article, a pair or a set, and says it need not be the unit of purchase or shipment. So 5,000 pairs of socks is a lot of 5,000: at General Level II the AQL calculator gives 200 pairs to inspect, and a pair with one faulty sock is one defective unit. Underwear in three-packs is different: 5,000 packs can be entered as 5,000, with 200 inspected, or as 15,000 garments, with 315 inspected, and the two are not equally strict, so agree the choice. Write the unit of inspection next to the AQL in your purchase order, and tell us if your retailer’s manual defines it.

Enter the figures the report states, then compare the plan the calculator gives with the plan the inspector used. Enter the quantity offered for inspection, not the order quantity: if 4,800 pieces of a 5,000-piece order were packed on the day, the lot is 4,800. That still falls in the 3,201 to 10,000 band, so at General Level II the plan is 200 garments, with 10 accepted and 11 rejected at AQL 2.5, and 14 accepted and 15 rejected at AQL 4.0. Then check four things in the report. The sample size matches the calculator’s, or the report names the arrow it followed. Majors and minors were each judged against their own accept number. If your order sets critical defects at zero, one critical defect led to a reject. And the report says whether it counted defects or defective garments, because the same accept numbers are stricter when every fault is counted. A pass recorded on fewer garments than the plan, or on more defects than the accept number, does not meet the plan in your purchase order.

Because a fixed percentage checks too few garments on a small lot and far more than the table asks for on a large one. The ISO 2859-1 sample grows much more slowly than the lot: at General Level II it is 20 garments for a lot of 120, where 10% would be 12; 50 for a lot of 300, where 10% would be 30; and 315 for a lot of 30,000, where 10% would be 3,000 (our arithmetic on the table’s figures). The table also comes with an accept and a reject number for each AQL, and the standard prints operating characteristic values that show how often a lot of a given quality passes. A 10% check has neither, so its pass or fail is a judgement made on the day, not a rule both sides agreed before production. If you want more garments inspected than Level II gives, choose Level III rather than a percentage: on 30,000 pieces it raises the sample from 315 to 500. Write the level and the AQLs into the purchase order.