ASVAB Score to IQ: What an AFQT Percentile Really Converts To

Search for an ASVAB-to-IQ chart and you will find dozens, each pairing military test scores with IQ figures, sometimes to a decimal point. None of them comes from the Defense Department. The number that decides whether someone can enlist, the AFQT, is a percentile rather than an IQ, and turning it into one rests on arithmetic that works only roughly. The same arithmetic settles one of the most repeated claims about military testing: that it is illegal to enlist anyone with an IQ below 83.

What the ASVAB and the AFQT actually are

The Armed Services Vocational Aptitude Battery, or ASVAB, is the test US military applicants take. Its subtests feed several scores used to match people to jobs, but one score decides basic eligibility: the Armed Forces Qualification Test score, or AFQT. Since January 1989 the AFQT has been built from four subtests, Arithmetic Reasoning, Mathematics Knowledge, Word Knowledge and Paragraph Comprehension. The two verbal subtests are combined into a verbal score that counts double, so the formula is AR + MK + 2(VE).

The result is reported as a percentile from 1 to 99. A percentile is a rank: an AFQT of 50 means scoring as well as or better than half of a reference group. That group comes from the 1997 Profile of American Youth, a joint Defense Department and Department of Labor study that tested about 6,000 people aged 18 to 23, found by screening more than 90,000 housing units. Its norms have been in use since July 2004, when they replaced norms from 1980. Scores are grouped into categories:

• Category I: 93rd to 99th percentile
• Category II: 65th to 92nd
• Category IIIA: 50th to 64th
• Category IIIB: 31st to 49th
• Category IVA: 21st to 30th
• Category IVB: 16th to 20th
• Category IVC: 10th to 15th
• Category V: 1st to 9th

Converting an AFQT percentile to an IQ-style number

Because the AFQT is a percentile, the only defensible conversion is the one that works for any percentile: find the matching point on a normal curve. On the mean-100, standard-deviation-15 scale most IQ tests use, that gives the equivalents below. Our IQ percentile calculator runs the same arithmetic in the other direction.

• AFQT 99: about 134.9
• AFQT 93, the bottom of Category I: about 122.1
• AFQT 65, the bottom of Category II: about 105.8
• AFQT 50: 100
• AFQT 31, the bottom of Category IIIB: about 92.6
• AFQT 21, the bottom of Category IVA: about 87.9
• AFQT 10, the lowest score policy allows: about 80.8

Treat these as rough equivalents, for three reasons. The reference groups differ: the AFQT ranks people against 18- to 23-year-olds tested in 1997, while an IQ test ranks them against its own age-matched standardisation sample, normed at a different time. The content differs: the AFQT is built entirely from arithmetic, mathematics knowledge, vocabulary and reading comprehension, without the nonverbal reasoning, working memory and processing-speed tasks a full intelligence battery includes, a distinction our note on achievement tests versus IQ tests explains. And every score carries measurement error, so one percentile is a range rather than a point.

Research does show the battery measures something close to general ability. A 2004 study in Psychological Science by Frey and Detterman extracted a general factor from the ASVAB scores of 917 young people in the 1979 National Longitudinal Survey of Youth and found it correlated .82 with their SAT scores. That supports treating the ASVAB as a strong measure of general ability. It does not supply a validated table from AFQT percentiles to IQ points, which is the same gap our look at Wonderlic score to IQ conversion found in another widely quoted formula.

The IQ 83 claim, checked against the rules

The claim that the US military is legally barred from enlisting anyone with an IQ below 83 has circulated for years through online lectures and video clips. No statute or regulation contains an IQ figure. The rules are written in AFQT percentiles:

• Defense Department policy: no one scoring below the 10th percentile, Category V, may be inducted or enlisted, a rule repeated in federal regulations
• Federal law, 10 U.S.C. § 520: since 22 December 2023, no more than 4% of a service's new active-duty enlistees in a fiscal year may score in Category IV, the 10th to 30th percentile; the Secretary of Defense may raise that to 20% after notifying Congress
• The same law: applicants without a high school diploma must score at the 31st percentile or above, unless a service needs them to meet its strength targets
• Defense Department benchmarks: at least 60% of recruits at the 50th percentile or above, and at least 90% high school diploma graduates

Convert the policy floor with the same normal-curve arithmetic and the 10th percentile corresponds to an IQ-style 80.8, not 83. An IQ of 83 sits at the 12.9th percentile, inside Category IVC, which the rules restrict but do not prohibit. The figure that does appear in academic writing is 80: in a 1997 paper in the journal Intelligence, the psychologist Linda Gottfredson described the legal minimum as the 10th percentile, about IQ 80, using the norms of that time.

For most applicants the real bar sits well above that floor. The Air Force and Marine Corps require an AFQT of 31 from high school graduates and 50 from GED holders, and the Coast Guard requires 32. The Army's enlistment regulation, updated in March 2026, lists scores of 10 to 15 as not currently accepted, with one narrow occupational exception. Actual recruits score higher still: in fiscal year 2019, 69% of new active-duty enlistees scored at the 50th percentile or above.

Where the lower categories come in

The Category IV rules have mattered in recent years. Facing recruiting shortfalls, the Army opened its Future Soldier Preparatory Course at Fort Jackson in August 2022; its academic track took applicants scoring 21 to 30 and aimed to raise them to 31 or above, and about 32,000 people had graduated from the course by January 2025. The Navy launched its own Future Sailor Preparatory Course on 10 April 2023 for applicants scoring 10 to 30, requiring a gain of at least 10 points. A May 2025 report by the Defense Department Inspector General found that the Army had used retest scores to move course trainees out of Category IV, and that without that reclassification it did not comply with the law or the department's policy.

The military has scored recruits on its own scales for more than 80 years, and the Army General Classification Test behind the most-shared chart of IQ by job dates from the Second World War, as our piece on average IQ by occupation shows. Lower-scoring recruits have been admitted deliberately before, too. Under Project 100,000, begun in 1966 under Defense Secretary Robert McNamara and ended in December 1971, the services took in men scoring between the 10th and 30th percentiles: more than 320,000 by 1971 according to a 1989 study for the Defense Department, and 341,127 by the count of a 2026 analysis. That 1989 study found the veterans later had less education, lower incomes and higher unemployment and divorce rates than non-veterans.

What an AFQT score cannot tell you

• It is not an IQ, and no official table turns it into one
• It ranks you against 18- to 23-year-olds tested in 1997, not against today's adults of your own age
• It leaves out the nonverbal reasoning, working memory and processing-speed tasks that IQ batteries measure
• It carries measurement error like any test score, so a few points either way are noise rather than a real difference

The short version: an ASVAB result converts to an IQ-style number only in the loose sense that any percentile does, the military itself makes no such conversion, and the rule behind the viral 83 is really a 10th-percentile floor on a 1997 youth sample. If you want a figure that is an IQ score in the ordinary sense, with its percentile reported beside it, our IQ test provides one.

Common questions