AFOQT Math Knowledge Practice Test (25 Questions, 22 Minutes)

A recall test wearing a maths test’s clothes

Twenty-two minutes for twenty-five questions is about fifty-three seconds each. That is enough to apply a rule you remember and nowhere near enough to derive one you have forgotten. The questions are short, the numbers are clean, and the failure mode is almost always the same: you knew this once.

That has a direct consequence for how to prepare. Working more practice questions is the obvious move and it is only half right — practice tells you which rules have gone, but re-learning them is a separate, faster exercise that most candidates skip in favour of doing more questions.

Build the formula sheet from memory

The single most efficient thing you can do for this subtest: sit down with a blank page and write out every formula and rule you expect to need, without looking anything up. The gaps in what you produce are precisely your study list. Repeat until the page comes out complete twice running.

This takes twenty minutes and reliably finds more lost points than an hour of practice questions, because practice questions only test the rules that happen to appear in them.

What actually shows up

• Linear equations and inequalities. Including the sign flip when you multiply or divide by a negative — the most common careless error on the subtest, and one that always has a matching wrong answer.
• Factoring. Difference of squares and simple trinomials. Recognising a² − b² = (a + b)(a − b) on sight turns a sixty-second question into a ten-second one.
• Exponent rules. Multiplying adds exponents, a power of a power multiplies them, a negative exponent is a reciprocal, anything to the zero is one.
• Geometry formulas. Area and circumference of a circle, areas of triangles and rectangles, volumes of cylinders and rectangular solids, and the interior-angle sum (n − 2) × 180°.
• Pythagorean triples. 3-4-5, 5-12-13, 8-15-17 and their multiples. Recognising 9-12-15 as a scaled 3-4-5 saves you squaring anything.
• Coordinate geometry. Slope between two points, and slope-intercept form. Inverting the slope fraction is a standard trap.
• Radicals and simplification. Pull out the largest perfect square: √72 is 6√2, and 3√8 is the same value but not fully simplified, so it appears as a wrong answer.
• Systems of equations. Usually solvable by adding or subtracting the two equations directly.

Back-solving is using the format, not cheating it

The options are given to you. For an equation with numeric answers, substituting the middle option often resolves the question faster than solving forwards, and tells you which direction to move if it is wrong. On a question asking which expression equals another, testing a single convenient value of x eliminates most options immediately.

This is a legitimate technique for a multiple-choice test under time pressure. Use it when the algebra looks long and the options look clean.

Practise the arithmetic you are not allowed to shortcut

No calculator, so mental arithmetic is part of the subtest whether or not it is being tested. Three habits earn back real time: doubling and halving to clear decimals (1,260 ÷ 3.5 becomes 2,520 ÷ 7), working percentages through ten percent, and estimating before computing so a wrong setup announces itself.

Pacing

Fifty-three seconds is a comfortable average and a tight limit. Most items resolve in twenty seconds if the rule is in memory, which banks time for the handful that need working. If an item is taking more than ninety seconds, you are almost certainly solving it the long way — mark it, move, and come back if the section allows.

There is no penalty for guessing, so fill in everything before the section closes. On this subtest a partly remembered rule is often enough to eliminate two options.

Why this one is worth the hours

Mathematics Knowledge appears across more of the composite structure than any other single subtest, which makes it unusually hard to waste time on. It is slower to move than Word Knowledge or Table Reading if your algebra has gone genuinely quiet, but the material is finite and it does come back.

We do not publish composite minimums, because they vary by commissioning source and we have not verified current figures against an official document. How the composites are built is explained on the scores and results page; confirm the numbers you need with your recruiter or cadre.

Rebuilding maths that has genuinely gone

If the formula-sheet exercise produces a mostly blank page rather than a few gaps, you are in a different situation from the candidate who just needs a refresher, and the plan should be different too. Working practice questions will mostly demonstrate that you cannot do them.

Rebuild in the order the topics depend on each other: arithmetic with fractions and negatives first, then manipulating linear equations, then factoring, then geometry formulas, then coordinate geometry. Skipping ahead to the topics that appear most often is tempting and does not work, because the later material assumes the earlier.

Budget properly for this. It is the one part of AFOQT preparation that genuinely rewards eight to twelve weeks rather than four, and it is the reason our study-duration guide treats a rusty mathematics foundation as the main argument for a longer runway.

Recall, not reasoning

The most useful thing to understand about this subtest is how it differs from Arithmetic Reasoning, because candidates who conflate the two study the wrong thing. Arithmetic Reasoning gives you a situation and asks you to model it. This gives you a piece of mathematics and asks whether you remember the rule.

That means the failure mode is decay rather than difficulty. People who did well in school mathematics routinely find their algebra has quietly gone — exponent rules, factorising, the relationships in a right triangle — and the fix is revision of specific rules rather than practice at problem-solving.

It also means improvement here is fast and measurable, which makes it a good early target for anyone whose diagnostic put it near the bottom.