ASVAB (Armed Services Vocational Aptitude Battery) Exam

The AFQT — the enlistment-gate number — is computed from four subtests only: Arithmetic Reasoning, Mathematics Knowledge, Paragraph Comprehension, and Word Knowledge, reported as a 1-99 percentile and sorted into categories from I (93-99) down to V (1-9). The remaining subtests — General Science, Electronics Information, Auto and Shop Information, Mechanical Comprehension, and Assembling Objects — feed the composite line scores each branch uses for job qualification.

The practice bank here mirrors that structure: the AFQT math and verbal areas carry the most weight, the science and technical area covers the job-qualifying subtests, and a small strategy area drills the scoring and retest rules that shape enlistment timelines.

The ASVAB costs nothing to take. Testing is arranged through a recruiter and administered by the Department of Defense at one of 65 Military Entrance Processing Stations across the United States and Puerto Rico, or at a satellite Military Entrance Test site closer to home. Bring valid identification: a candidate without it, or one who arrives late, is turned away and rescheduled rather than tested.

The real cost is the retest clock. A candidate who wants a higher score must wait one month before a second attempt, one more month before a third, and six months before every attempt after that. A large jump is checked rather than accepted: an AFQT gain of 20 or more points within six months triggers a required Confirmation Test, and the confirmation result governs. A test invalidated for administrative reasons is treated differently from a retake, and the retest policy sets out those cases. Because the waits compound, the first attempt should be taken as the real one — a practice bank and the score categories above are the preparation, not a throwaway first sitting.

Nearly everyone takes the computer-adaptive CAT-ASVAB: each subtest starts at a middle difficulty and selects a harder item after a correct answer and an easier one after a miss, so two candidates never see the same test. The scored question counts and time limits per subtest are fixed:

The four AFQT subtests — AR, MK, PC and WK — carry the enlistment decision, and the time limits show where the pressure sits: 55 minutes for 15 arithmetic problems is generous, 9 minutes for 15 vocabulary items is not. Unfinished items are scored as random guesses, so a subtest should never be left with questions unanswered. The paper version, given at only a handful of test sites, combines Auto and Shop Information into one subtest and drops Assembling Objects.

Scores come back as standard scores normed to 18-to-23-year-olds — about half of that group scores 50 or above and about one in six scores 60 or above — and the AFQT as a percentile from 1 to 99 against the 1997 norming sample, sorted into categories from I (93–99) to V (1–9). Each service branch sets its own minimum AFQT for enlistment and publishes it on its own site; the ASVAB program does not set one.

The ASVAB is unusual among the exams on this site in that "how hard is it" has no fixed answer, because there is nothing to pass. The battery does not return a pass or fail. It returns a set of standard scores and one percentile — the AFQT — and what that percentile has to clear is set by the branch you are talking to, not by the test. Two candidates can sit the same battery, post the same numbers, and walk out with one eligible and one not, because they applied to different services. The difficulty, then, is not the questions. It is the gap between the number you can produce and the number your target requires — and the fact that you get very few clean attempts at closing it.

What the score actually means

Every subtest is reported as a standard score normed against 18-to-23-year-olds: roughly half of that reference group scores 50 or above on any subtest, and roughly one in six scores 60 or above. So a 50 is average by construction and a 60 is a genuinely strong result, not a "C".

The AFQT is computed from four of those subtests — Arithmetic Reasoning, Mathematics Knowledge, Paragraph Comprehension and Word Knowledge — and reported as a percentile from 1 to 99 against the 1997 norming sample. The percentile is then sorted into categories: I is 93–99, II is 65–92, IIIA is 50–64, IIIB is 31–49, IVA is 21–30, and so on down to V at 1–9. Those category lines are the ones recruiters talk in, and the practical consequence is that a single percentile point can matter a great deal at a boundary (49 versus 50) and almost not at all in the middle of a band.

The remaining subtests — General Science, Electronics Information, Auto Information, Shop Information, Mechanical Comprehension and Assembling Objects — never touch the AFQT. They feed the line-score composites each branch uses to qualify you for specific occupations. This split is the single most important thing to understand before you study: the AFQT four decide whether, the rest decide what.

Why the adaptive format changes how you should take it

Almost everyone sits the computer-adaptive CAT-ASVAB. Each subtest opens at middling difficulty and adjusts as you go — a correct answer draws a harder item, a miss draws an easier one — so no two candidates see the same set. Three things follow from that mechanic, and none of them is intuitive.

First, there is no going back. An adaptive test commits to a path with each answer, so the habit of skipping hard items and returning later does not exist here. You answer what is in front of you or you do not.

Second, an unfinished subtest is scored as if the remaining items were random guesses. A subtest left with three items unanswered is scored as three coin-flips, not three blanks. On a time-tight subtest this means a fast, deliberate guess on the last item is always better than the clock running out on it.

Third, the time limits are set so that finishing is expected — but the pressure is wildly uneven across subtests. Word Knowledge gives you 15 items in 9 minutes: 36 seconds each. Arithmetic Reasoning gives you 15 items in 55 minutes: well over three minutes each. Candidates who practise against a single generic clock arrive tuned wrongly for both.

Where the pressure sits, subtest by subtest

The scored counts and limits are fixed, and they tell you where to build speed and where to build care:

• Word Knowledge — 15 items, 9 minutes. The tightest clock on the battery, and an AFQT subtest. This is pure recall; either you know the word or you do not, and dwelling costs more than guessing.
• Paragraph Comprehension — 10 items, 27 minutes. AFQT. Generous time, so losses here are comprehension losses, not pacing losses.
• Arithmetic Reasoning — 15 items, 55 minutes. AFQT. The most time per item on the battery; the failure mode is arithmetic slips under low pressure, not running out.
• Mathematics Knowledge — 15 items, 31 minutes. AFQT. Two minutes an item — comfortable if the algebra is fluent, punishing if you are re-deriving formulas.
• General Science — 15 items, 12 minutes. Line scores only. Fast recall.
• Electronics Information — 15 items, 10 minutes. Line scores. Forty seconds an item.
• Auto Information — 10 items, 7 minutes; Shop Information — 10 items, 6 minutes. Line scores. The shortest subtests on the battery.
• Mechanical Comprehension — 15 items, 22 minutes. Line scores.
• Assembling Objects — 15 items, 18 minutes. Line scores; spatial, and the one subtest people cannot cram for.

Our bank weights the four AFQT subtests and the technical subtests roughly evenly, which is deliberate: the AFQT decides eligibility, but a candidate who neglects Electronics and Mechanical Comprehension can enlist and then find the occupations they wanted closed to them.

The paper version is a different test

The paper-and-pencil ASVAB survives at only a handful of Military Entrance Test sites. If you are sent to one, two things change: Auto Information and Shop Information merge into a single Auto & Shop subtest, and Assembling Objects is not given at all. Prep material built for the CAT-ASVAB will describe a subtest you will not see and split a subtest you will see combined — confirm which version your test site administers before you build a study schedule around the table above.

Retesting is where the real cost lives

There is no fee, and that fact is responsible for a great many under-prepared first sittings. What the ASVAB charges instead is time. A candidate who wants a higher score must wait one month for a second attempt, one month more for a third, and six months for every attempt after that. Three sittings put you two months out; a fourth puts you eight months out.

And a large improvement is not simply accepted. An AFQT gain of 20 or more percentile points within six months triggers a mandatory Confirmation Test, and the confirmation result is the one that stands. A candidate who genuinely improves by 20 points has nothing to fear from that; a candidate hoping a lucky day will carry them should understand that a lucky day is precisely what the confirmation rule exists to catch. Plan to arrive at your first sitting already scoring where you need to be on full-length practice, because the sittings that follow are expensive in the only currency the ASVAB uses.

A realistic plan

1. Find your target before you open a book. Ask the recruiter for the AFQT minimum for the branch you want and the line-score minimums for the occupations you want. Studying without those two numbers is studying without a destination.
2. Diagnose against the real clocks. One full CAT-format practice battery, subtest by subtest, with the time limits above. You are looking for two different things: AFQT subtests where you are below target, and technical subtests where you are far below the occupations you want.
3. Spend AFQT time on Word Knowledge and Mathematics Knowledge first. They are the two AFQT subtests where preparation moves the score fastest — vocabulary is recall, and MK is formula fluency. Arithmetic Reasoning has so much time per item that gains there come from accuracy, not speed.
4. Drill Word Knowledge under its own 36-second clock. It is the subtest most likely to be left unfinished, and unfinished items are scored as guesses.
5. Do not neglect the technical subtests if the job matters to you. Enlistment is the gate; the line scores are the building.
6. Sit once, ready. The month-then-month-then-six-months schedule makes the first attempt the one that counts.