The RACGP AKT contains 150 single-best-answer and modified extended-matching questions. Its base duration is three and a half hours, but a universal additional 30-minute allowance is applied to every candidate, giving all candidates a full four hours to complete the paper. That works out to an effective average of approximately 96 seconds per question, a genuinely workable pace for most candidates provided it is managed deliberately rather than left to chance.
Why the average matters less than how it is spent
Ninety-six seconds per question, treated as a strict per-item limit, is the wrong way to think about pacing this exam. Some questions, particularly straightforward single-best-answer items testing recognition of a well-known presentation, can be answered comfortably in well under a minute. Others, particularly extended-matching questions requiring genuine comparison between several plausible management options, deserve considerably more time. The average exists to be spent unevenly across the paper, not enforced uniformly on every question.
A three-tier approach to question difficulty
It helps to sort questions, as they are encountered, into three rough categories rather than treating the paper as uniformly difficult. Straightforward applied-knowledge items, where the correct answer is essentially clear once the stem has been read properly, should be answered promptly, banking time for later in the paper. Questions requiring genuine comparison between two plausible management options deserve a fuller, but still bounded, amount of consideration, since these are usually where real marks are won or lost. And items genuinely worth flagging for later review, where the first read leaves real uncertainty, should be marked and moved past quickly rather than allowed to consume disproportionate time on a first pass.
Section checkpoints rather than a running clock
Watching the clock continuously tends to increase anxiety without improving pacing discipline. A more effective approach uses fixed checkpoints, roughly every 30 questions, against a rough time budget: by question 30, a candidate should have used somewhere in the region of 45 to 50 minutes; by question 60, roughly 90 to 100 minutes; and so on through the paper. Checking progress at these intervals, rather than continuously, allows genuine drift to be caught and corrected early, while leaving enough uninterrupted attention for the questions themselves in between checkpoints.
The real cost of protecting a single mark
It is worth being explicit about the opportunity cost of spending several minutes defending one difficult answer. A candidate who spends four extra minutes deliberating over a single uncertain question, in the hope of squeezing out one additional mark, is implicitly trading that time against several other questions later in the paper that could each have been answered correctly with a normal amount of consideration. Given that every question on the AKT carries equal weight, this trade is very rarely a good one, and the discipline to move on from a genuinely uncertain question, having made the best available judgement, is a skill worth deliberately practising rather than assuming will simply happen under pressure.
Building pacing through progressively timed practice
Realistic pacing under genuine time pressure is not something most candidates arrive at naturally; it needs building through staged practice. Timed blocks of 25 questions establish the basic rhythm without the pressure of a long session. Blocks of 50 questions introduce a meaningful degree of sustained concentration. Blocks of 75 questions begin to test whether pacing holds up as fatigue starts to set in. And full 150-question, four-hour simulations, attempted more than once in the final stretch of preparation, are the only way to genuinely confirm that a candidate's pacing strategy survives contact with the complete paper.
Detecting fatigue through quartile analysis
A particularly useful piece of self-analysis, once full-length practice sessions are underway, is comparing accuracy across the four quartiles of the paper, roughly questions 1 to 37, 38 to 75, 76 to 112, and 113 to 150. A meaningful drop in the final quartile, compared with the first, is a strong signal of fatigue rather than knowledge, and points towards a different kind of preparation, building endurance and concentration stamina, rather than simply learning more content.
An exam-day pacing plan
On the day itself, a workable approach combines the checkpoint system above with a clear, pre-decided rule for what happens when a checkpoint shows the pace has slipped: moving more briskly through straightforward questions to recover lost time, rather than attempting to speed up uniformly across every remaining question, tends to be the more reliable correction, since it protects the time still available for the harder items yet to come.
Why the 30-minute universal allowance changes the maths less than it feels like it should
Because the additional 30 minutes is applied universally, to every candidate regardless of individual circumstances, it does not represent extra breathing room relative to other candidates in the way that supplementary time for a specific individual need might. It simply resets the effective baseline pace for the whole cohort from around 84 seconds per question, under a strict three-and-a-half-hour reading of the paper, to roughly 96 seconds. Candidates sometimes treat the allowance as a buffer to be held in reserve for emergencies late in the paper, which is a reasonable instinct, but the safer approach is to build the full four hours into pacing calculations from the outset, rather than mentally reserving the final 30 minutes and then pacing the first 150 minutes as though working to the shorter, stricter figure.
What to do if a checkpoint reveals a serious problem
Occasionally a checkpoint reveals not minor drift but a genuine pacing crisis, for instance reaching question 75 with less than half the total time remaining. In this situation, the correct response is not panic-driven rapid guessing through the remainder of the paper, but a deliberate, conscious shift: spend noticeably less time on questions that read as straightforward, protect slightly more time for the small number that are genuinely difficult, and accept that a few uncertain answers, chosen quickly using the elimination method rather than left to the very end, will need to stand. Practising this kind of recovery specifically, deliberately inducing a pacing crisis during a mock and rehearsing the recovery, is a worthwhile exercise precisely because it is unpleasant to do and therefore rarely practised voluntarily.