If your ACT math plan is "about a minute a question," you are pacing for a test that no longer exists.
That rule came from the old format. Sixty questions, sixty minutes, easy to remember. It is also no longer true. The enhanced ACT hands you 45 questions and 50 minutes, which works out at 67 seconds each rather than 60. Seven seconds sounds like nothing. Multiply it across the whole section and you are looking at more than five minutes.
Most pacing advice you will find still assumes the old format. Here is what the test actually looks like now, taken from ACT's own published specifications rather than from anyone's summary of them.
What the section is now
ACT's design framework lays the two versions side by side:
- Legacy math: 60 questions, 60 minutes, 60 seconds per item, five answer choices.
- Enhanced math: 45 questions, 50 minutes, 67 seconds per item, four answer choices.
Two other changes matter as much as the clock. The composite score is now built from English, math and reading only, with science optional, so math carries a third of your composite instead of a quarter. And dropping from five answer choices to four quietly improves every guess you make.
So the section got shorter, slower per question, and more important. That combination should change how you run it.
The content shifted too
The blueprint moved along with the timing, and it moved in one direction: harder.
ACT sorts math items into two buckets. "Preparing for Higher Math" covers number and quantity, algebra, functions, geometry, and statistics and probability. "Integrating Essential Skills" covers the more elementary material. On the legacy test, the higher-maths bucket was 60 percent of the section. On the enhanced test it is 80 percent, roughly 33 of the 41 scored questions.
Within that, algebra, functions and geometry each take something like 17 to 20 percent, with number and quantity around a tenth. So a rough expectation for any given form is seven or eight questions each on algebra, functions and geometry, four or five on number and quantity, and the balance on statistics and probability.
Fewer questions, a higher proportion of them advanced. Another reason a flat pace is the wrong instinct.
Four of your questions do not count
This is the part almost nobody mentions, and it has a real strategic consequence.
Of the 45 questions, only 41 are scored. The other four are embedded field-test items, questions ACT is trialling for future exams. You can confirm the 41 independently: ACT's scoring documentation lists the highest possible raw score for mathematics as 41.
You cannot identify them. ACT's framework states that field-test items are "placed strategically throughout the section tests, blending seamlessly with operational items," and that in math specifically they "will be inserted throughout the test as discrete items." They are deliberately camouflaged, and they are timed exactly like everything else.
The consequence is not that you should try to spot them. It is that when a question feels strange, badly worded, or unlike anything you have practised, there is a real chance it is one of the four that will never touch your score. Digging in for three minutes on principle is a bad bet twice over: you might lose the time for nothing at all.
The questions get harder as you go, officially
ACT's math content blueprint describes the items as "ordered in increasing difficulty." That is not a rumour from a prep forum. It is the design.
Which means a flat pace is the wrong model. Spending 67 seconds on question 3 and 67 seconds on question 43 treats two very different problems as equivalent, and the early ones are where you can bank time cheaply.
A budget shaped like the test
Split the 45 questions into thirds and give each third a different rate.
- Questions 1 to 15: 40 seconds each. Ten minutes total.
- Questions 16 to 30: 65 seconds each. About sixteen minutes.
- Questions 31 to 45: 90 seconds each. About twenty-two minutes.
That comes to roughly 49 minutes, leaving about a minute of slack. The whole point is that the last third gets nearly a minute more per question than the average, and it is paid for entirely by moving briskly through the first third.
Two checkpoints are enough to run this without watching the clock constantly. You want to be finishing question 15 at around the ten-minute mark, and question 30 at around twenty-six minutes. If you hit both, the hard third has all the time it needs. If you are behind at question 15, you are not behind on the hard questions yet, and you can still fix it.
The mistake ACT actually watched students make
ACT ran cognitive labs while designing this test, and the observations in their report are more useful than most strategy advice, because they are descriptions of real behaviour rather than theories about it.
In one study, students spent more than four minutes on a single graphing item before guessing an answer anyway, without having used their graphing calculators. Four minutes, and then a guess. That is roughly four questions' worth of budget spent to arrive exactly where they would have arrived immediately.
The finding underneath it is starker. Of the low-scoring students observed, none skipped or guessed and moved on. Not some. None. They knew, when surveyed, that you are supposed to move on. Under real conditions, not one of them did it.
That gap between knowing the strategy and executing it is the actual thing to practise. It is not knowledge. It is the willingness to abandon a question that is beating you, which is uncomfortable in a way that no amount of understanding the advice fixes. If you use a graphing tool as part of your prep, practise the decision as well as the mechanics: plotting a function to find where curves cross should take seconds, and if it is taking minutes, that is the signal to guess and move.
Guessing got better
The drop from five answer choices to four moved a blind guess from a one in five shot to a one in four. Across a handful of questions that is a real difference.
And ACT does not deduct anything for a wrong answer. Their scoring documentation is explicit that raw scores are simply the number of correct answers, with no penalty for guessing. A blank and a wrong answer score identically, which makes leaving anything blank strictly worse than filling something in.
So the rule is mechanical. When a question hits your time limit, pick an answer, mark it, and move. You can come back if the clock allows. You have already banked the 25 percent either way.
Practise the clock, not just the maths
Pacing is a separate skill from doing the problems, and it is the one that usually goes untrained. Doing a problem set at your kitchen table with no timer builds the maths and none of the judgment.
Run at least some of your practice under the real constraint: 45 questions, 50 minutes, checkpoints at 15 and 30. You will discover very quickly whether your problem is the mathematics or the clock, and those need completely different work. If it turns out the material is solid at home and evaporates under time, that is a different mechanism entirely, and worth diagnosing properly rather than answering with more practice sets.
One more thing worth borrowing from the digital SAT side of test prep, where the calculator is always available: the tool is not the strategy. Knowing when to skip the calculator entirely saves more time on that test than fluency with it does, and the same judgment applies here. Your 67 seconds go to thinking, not to typing.