The RACGP's 2026.1 exam report for the Applied Knowledge Test recorded 1,117 candidates, a mean score of 75.79%, a cut score of 67.11%, an overall pass rate of 80.57%, and a reliability coefficient of 0.91. These figures are worth reading properly rather than skimming, because each one answers a different question, and conflating them is where candidates most often go wrong in how they interpret their own readiness.
How the pass mark is actually set
The RACGP determines the pass mark for each AKT sitting using a modified Angoff process. In practice, this means a panel of experienced GP examiners reviews the question paper and estimates, question by question, the probability that a just-competent, newly fellowed GP would answer it correctly. The resulting cut score reflects the specific difficulty of that sitting's paper, not a fixed percentage the college has decided candidates must clear. The RACGP does not operate quotas; there is no predetermined number or proportion of candidates who are permitted to pass, and the college has said so explicitly in its own published guidance.
Four figures, four different meanings
It is worth being precise about what each number in the report actually tells you. The pass mark, 67.11% in this sitting, is the threshold a candidate needed to clear, set specifically for this paper's difficulty. The pass rate, 80.57%, is simply the proportion of candidates who cleared that threshold; it describes the cohort's collective performance, not any individual's chance of passing. The mean candidate score, 75.79%, describes the average performance across the whole cohort, and sits comfortably above the cut score, which is a reasonably normal pattern for this exam. And the reliability coefficient, 0.91, is a psychometric measure of how consistently the exam paper actually discriminated between stronger and weaker candidates; a figure this high suggests the paper was measuring something real and stable, rather than being unusually noisy or arbitrary in how it separated passing from failing performance.
Why 67.11% is not a safe target for a future sitting
The single most consequential mistake a candidate can make with this report is treating 67.11% as a number to aim for in their own preparation. Because the cut score is set fresh for every sitting, based on that specific paper's difficulty profile, there is no guarantee the next sitting's threshold will land anywhere near this figure. A paper drawn from harder or more discriminating questions could produce a lower cut score for the same underlying standard, and a more straightforward paper could produce a higher one. Preparing to just clear the most recently published cut score leaves no margin at all if the next paper happens to set a marginally higher bar, and no margin is a genuinely risky position to be sitting an exam in.
What a sensible margin actually looks like
Rather than targeting the published cut score directly, a more defensible approach is to maintain a consistent buffer above it in fresh, mixed practice, ideally reflecting something closer to the reported mean score of 75.79% or higher, sustained across genuinely unseen questions rather than familiar ones. That margin exists specifically to absorb the sitting-to-sitting variation in where the actual cut score lands, along with the ordinary performance variability every candidate experiences between practice and the pressure of the real exam room.
Translating the report into preparation priorities
The report is more useful as a diagnostic input than as a headline statistic to quote back. A high reliability coefficient suggests that genuine differences in preparation quality are being captured accurately by this kind of exam, which is a reason to trust that consistent, broad-based practice really does translate into a better outcome rather than being at the mercy of an unpredictable paper. The gap between the mean score and the cut score, roughly nine percentage points in this sitting, is a reasonable proxy for the kind of buffer a well-prepared candidate should be aiming to build into their own practice performance well before exam day, rather than discovering on the morning of the sitting whether they have enough margin.
Common ways candidates misread this kind of report
Beyond treating the cut score as a target, there are a handful of other misreadings worth naming directly. Some candidates assume a high overall pass rate means the exam was somehow easier that sitting; in fact, a high pass rate combined with a high mean score is at least as likely to reflect a well-prepared cohort as an easier paper, and the reliability figure gives no direct information either way about paper difficulty in isolation. Others assume that because the college does not operate quotas, pass rates should be stable from sitting to sitting; in practice, cohort composition, the proportion of first-time versus repeat candidates, and the specific difficulty profile of each paper all vary, and the pass rate is simply where those factors happen to land for a given cycle. And some candidates read a strong mean score as reassurance that the exam is broadly forgiving; a mean of 75.79% sitting comfortably above a cut score of 67.11% says more about the shape of the score distribution for that particular cohort than it does about how much margin any individual candidate personally needs.
Why this report should shape practice conditions, not just content review
It is worth treating the report as an input into how practice is structured, not only what is studied. A reliability coefficient of 0.91 is high enough to suggest the paper is discriminating genuine differences in preparation reasonably consistently, which is itself a reason to trust that disciplined, broad-based practice under realistic conditions, rather than last-minute cramming or narrow question-spotting, is likely to be rewarded by this kind of exam. Candidates building their final weeks of preparation around full-length, timed, mixed-topic practice, with a genuine margin over the most recently published cut score, are responding to what this report actually implies about how the exam behaves, rather than simply reacting to the headline percentage.
A note on reproducing RACGP material
The figures cited in this article are drawn from the RACGP's own published exam report and are presented here as factual reference points. RACGP charts, tables and detailed report content are published under the college's own licence terms, and readers producing their own materials based on RACGP publications should check the current licence conditions on the RACGP's site before reproducing or adapting any of the college's original report material, rather than assuming public availability implies unrestricted reuse.