AFQT Explained: How Your Score Is Calculated and What Counts as Competitive

Last updated June 10, 2026

The AFQT — the Armed Forces Qualification Test — is the single number that decides whether the military will accept you at all. It is reported as a percentile from 1 to 99, and every branch sets a minimum AFQT you have to clear before any other conversation can happen. Your line scores decide which specific jobs you can do; your AFQT decides whether you are in the door in the first place.

Despite the name, the AFQT is not a separate test. It is a composite built from just four of your ASVAB subtests using the formula AFQT = 2 × VE + AR + MK, where VE (Verbal Expression) is your Word Knowledge plus Paragraph Comprehension, AR is Arithmetic Reasoning, and MK is Mathematics Knowledge. The other five ASVAB subtests do not count toward your AFQT at all — they only feed the line scores that gate individual jobs.

The most important thing to understand is that the resulting number is a percentile, not a percent-correct. An AFQT of 50 does not mean you got half the questions right; it means you scored as well as or better than about 50 percent of a nationally representative reference group of 18-to-23-year-olds. And there is a hard floor: by law, anyone scoring in Category V (the 1st through 9th percentile) is ineligible to enlist, no matter what.

What the AFQT actually measures

The ASVAB battery has nine subtests, but the AFQT only uses four of them. Verbal Expression (VE) bundles Word Knowledge and Paragraph Comprehension into one number — VE = WK + PC — and that verbal half is weighted double in the formula. The math half is Arithmetic Reasoning (AR), which is word problems, plus Mathematics Knowledge (MK), which is more abstract algebra and geometry.

Put together, the AFQT composite is 2 × VE + AR + MK. Because VE is doubled and is itself the sum of two subtests, reading and vocabulary carry a lot of the weight: WK and PC each effectively count twice, while AR and MK each count once. That is why applicants who are strong in math but weak in reading are sometimes surprised that their AFQT does not climb as fast as they expect — half of the composite is verbal.

The raw composite ranges from 0 to 400. It is the four subtest standard scores, combined by that formula, before anything is converted to a percentile. The 0–400 number is an intermediate step; the percentile is what the military and your recruiter actually use.

Percentile, not percentage

This is the part applicants most often get wrong. The AFQT percentile is a ranking against a fixed reference population — the youth who took the ASVAB during the Profile of American Youth 1997 (PAY97) study, which remains the official norming baseline. A percentile of 65 means you performed as well as or better than roughly 65 percent of that reference group. It says nothing about how many questions you answered correctly.

A practical consequence: percentile scores are not evenly spaced. The middle of the distribution is crowded, so a few extra correct answers near the average can swing your percentile by several points, while the same gain out at the high end barely moves it. Most people cluster around the middle, which is exactly why getting from a 45 to a 55 can feel achievable with focused study, but pushing from a 90 to a 95 is genuinely hard.

Because it is a relative ranking, your AFQT does not change based on which branch you apply to or which year you tested — it is benchmarked against the same reference population every time. What differs by branch is the minimum percentile each one requires, not how your percentile is computed.

The six AFQT categories

The military groups AFQT percentiles into six categories. These categories are the shorthand recruiters and policy use to talk about eligibility, because they line up with what jobs and incentives tend to open up. premil reports your estimated category alongside your percentile so you can place yourself on this scale at a glance.

Category I — 93rd to 99th percentile. Top tier. The widest possible eligibility; essentially every job is on the table from an AFQT standpoint.
Category II — 65th to 92nd percentile. Above average. Eligible for most jobs and the strongest position for competitive, score-gated specialties.
Category IIIA — 50th to 64th percentile. Solidly average. Still eligible for most jobs; this is the band most branches treat as fully competitive.
Category IIIB — 31st to 49th percentile. Below the median. You can enlist, but some jobs and some incentives start to close off, and a few branches set their floor above this range.
Category IV — 10th to 30th percentile. Limited eligibility. Enlistment is possible but tightly capped and restricted; far fewer jobs are available, and additional requirements often apply.
Category V — 1st to 9th percentile. Not eligible. Enlistment is barred by statute (10 U.S.C. § 520), so this is an automatic stop regardless of any other qualification.

What counts as competitive

There is no single national "passing" AFQT — each branch sets its own minimum, and those minimums move with recruiting needs. As a rule of thumb, scoring at the 50th percentile or higher (Category IIIA and up) puts you in the band most branches treat as fully eligible and keeps the widest set of options open. Below the median, you can often still enlist, but your choices narrow and some incentives disappear.

It is worth being clear about what a high AFQT does and does not do. A strong AFQT clears the enlistment gate comfortably and signals to a recruiter that you can handle most training pipelines, but it does not by itself guarantee a specific job. Individual jobs are gated by line scores — different composites of the subtests — so two applicants with the same AFQT can qualify for very different job menus depending on how their subtest scores are distributed. Think of the AFQT as the front door and the line scores as the keys to specific rooms.

Because branch minimums and category caps shift with policy and recruiting demand, treat any specific cut score you read online as general orientation only and confirm the current number with a recruiter. premil deliberately does not publish branch minimum cut scores as if they were fixed facts; what it gives you is your own estimated standing on this scale.

How premil estimates your AFQT (and why it is an estimate)

When you take a practice ASVAB in premil, it computes your AFQT composite the official way — 2 × VE + AR + MK — and then converts that 0–400 composite into a percentile. The conversion uses a statistical model of the PAY97 reference distribution: the composite is treated as approximately normal with a mean around 300 and a standard deviation of about 31.6, and the percentile is read off the normal curve. The result is then clamped to the 1–99 range that AFQT percentiles are always reported in.

This is an honest approximation, not the exact government answer. The Department of Defense publishes the real ASVAB only as an interactive computer-adaptive test, and the precise PAY97 conversion tables that turn a composite into an official percentile are not publicly available. premil uses a defensible normal-distribution model instead, which lands within roughly a couple of percentile points of independently reported benchmarks. So your premil number is a reliable ballpark for planning — good enough to know whether you are comfortably clear, borderline, or short — but the official percentile comes only from the real ASVAB administered at a test site.

The practical takeaway: use your premil estimate to decide where to focus. If you are sitting near a category boundary, the four AFQT subtests — Word Knowledge, Paragraph Comprehension, Arithmetic Reasoning, and Mathematics Knowledge — are exactly where extra study pays off twice, because they drive your AFQT and also feed many of the line scores that decide your jobs.