Skip to content
ASdrawn

AQL sampling plan calculator

A lot of 500 at general level II and an AQL of 1.0 means: inspect 50 items, accept the lot on 1 nonconforming or fewer, reject it on 2 or more. Enter your lot size, inspection level and AQL. The calculator reads the code letter, follows any arrow in the master table to the plan actually used, and draws the plan’s operating characteristic: how often a lot of any quality is accepted. The tables are MIL-STD-105E’s, the scheme ANSI/ASQ Z1.4 and ISO 2859-1 carry on.

Inspection
01234d05101520050100
Sample size n
Accept on Ac or fewer
Reject on Re or more
Code letter (Table I)
H
Plan read from letter
H
Accepted at the AQL
91.1%
Accepted 10 % of the time at
7.56%
How it’s calculated

Table I turns the lot size and the inspection level into a code letter. The master table for the inspection in force turns the letter and the AQL into the plan: n items, accept on Ac or fewer nonconforming, reject on Re or more. Where the table has an arrow, the plan is the first one below (↓) or above (↑) in the same AQL column, with that row’s sample size. The curve is the probability of acceptance, P(d ≤ Ac): binomial in n and p for percent nonconforming, Poisson with mean n·p/100 for nonconformities per 100 items.

Sample size code letters

Lot sizeS-1S-2S-3S-4IIIIII
2–8AAAAAAB
9–15AAAAABC
16–25AABBBCD
26–50ABBCCDE
51–90BBCCCEF
91–150BBCDDFG
151–280BCDEEGH
281–500BCDEFHJ
501–1200CCEFGJK
1201–3200CDEGHKL
3201–10000CDFGJLM
10001–35000CDFHKMN
35001–150000DEGJLNP
150001–500000DEGJMPQ
500001+DEHKNQR

Single sampling plans, normal inspection (Ac/Re)

Lettern0.0100.0150.0250.0400.0650.100.150.250.400.651.01.52.54.06.510152540651001502504006501000
A2↓↓↓↓↓↓↓↓↓↓↓↓↓↓0/1↓↓1/22/33/45/67/810/1114/1521/2230/31
B3↓↓↓↓↓↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/2230/3144/45
C5↓↓↓↓↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/2230/3144/45↑
D8↓↓↓↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/2230/3144/45↑↑
E13↓↓↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/2230/3144/45↑↑↑
F20↓↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑
G32↓↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑
H50↓↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑
J80↓↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑
K125↓↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑
L200↓↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑
M315↓↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑↑
N500↓↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑↑↑
P800↓0/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑↑↑↑
Q12500/1↑↓1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑
R2000↑↑1/22/33/45/67/810/1114/1521/22↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑

↓ use the first plan below, ↑ the first plan above, in the same AQL column, with that row’s sample size. Where the sample reaches the lot size, inspect every item.

Switching between normal, tightened and reduced

In MIL-STD-105E inspection starts normal, and each class of nonconformity switches on its own. The counts are of lots on their first inspection; a resubmitted lot does not count.

  • Normal to tightened: 2 lots rejected within 5 or fewer consecutive lots.
  • Tightened to normal: 5 consecutive lots accepted.
  • Normal to reduced, all of: the last 10 lots on normal inspection were accepted, the nonconforming found in their samples total no more than the standard’s limit number for them (its Table VIII, not reproduced here), production is steady, and reduced inspection is judged worthwhile.
  • Reduced to normal, any of: a lot rejected, a lot accepted in the reduced plan’s gap, irregular or delayed production, or other reason.
  • Stop: once the lots rejected during a spell of tightened inspection reach 5, acceptance under the scheme stops until the cause is corrected; it then restarts on tightened.

ISO 2859-1:1999 replaces the 10-lot limit-number test with a switching score that grows with each accepted lot, and its reduced plans differ (below).

MIL-STD-105E, ANSI/ASQ Z1.4 and ISO 2859-1

The tables here are MIL-STD-105E’s (1989), a public-domain US defence standard, cancelled in 1995 in favour of the industry’s ANSI/ASQ Z1.4-2003 (R2018), which carried its scheme on. ISO 2859-1 is the international scheme from the same line. Their code letters and single sampling plans for normal inspection agree cell for cell. Differences found against ISO 2859-1:1999:

  • Reduced inspection: the same sample sizes, but every plan has Re = Ac + 1 (for example 1/2, 2/3, 5/6, 6/7 where MIL-STD-105E has 0/2, 1/3, 2/5, 3/6), so there is no gap and no accept-then-revert result.
  • Tightened inspection: code letter A at AQL 10 is 0/1 on 2 items; MIL-STD-105E sends it down an arrow to letter D, 1/2 on 8 items.
  • The vocabulary: nonconforming and nonconformity for defective and defect, acceptance quality limit for acceptable quality level.

Not checked: the ANSI/ASQ Z1.4 text itself (licensed) and ISO 2859-1:2026, the third edition, which rearranged its tables. Use the edition your contract or drawing names.

What the AQL does and does not promise

The AQL is the worst process average the scheme treats as acceptable over a run of lots, not a limit on any one lot. A lot exactly at the AQL is accepted most of the time (between about 87 % and 99 % on normal inspection, by plan), and a lot several times worse still passes now and then: read the curve at the quality you cannot ship. The switching rules are what protect you over a run, by tightening after rejections.

Sampling and a first article

A first article inspection checks every characteristic on a part from the first production run; it is not an acceptance sample. Sampling by AQL belongs to the lots that follow, where the contract or the customer’s quality clauses allow it, and where they name the level and the AQL.