This workflow is for RACGP registrars sitting the Applied Knowledge Test (AKT) and Key Feature Problem (KFP) who hold a small, finite set of official RACGP materials and want to use them as a calibration instrument rather than burning them as daily practice. It addresses the two written components. Its principal limitation is scarcity: there are very few official items, and the moment you use one for casual revision you can never again use it as a clean, unseen measure of where you stand.
What "RACGP Official Resources" actually are
There is no official RACGP mega-bank. The genuinely official material is a compact set of high-authority documents, summarised below (availability last checked 19 July 2026 — confirm on racgp.org.au, as pages are updated between diets).
| Item | What it is | Rough volume | Best single use |
|---|---|---|---|
| Sample AKT questions | Worked single-best-answer examples | small set | one-time format and standard calibration |
| Sample KFP questions | Worked KFP cases with scoring notes | a handful (around five) | one-time KFP scoring calibration |
| Curriculum and syllabus | The examinable blueprint | full | coverage mapping |
| Candidate guidelines | Rules, conduct, response format | full | condition-setting for mock sittings |
| Examiners' and exam reports | Standard-setting and common-error commentary | periodic | error-pattern insight |
There is no AI, no adaptivity and no large question volume here — and that is precisely the point. These materials are authoritative and unseen, which makes them a calibration gold-standard. A third-party bank gives you volume; the official set gives you a true north. Confuse the two and you waste the more valuable asset.
The exam you are calibrating against
Per racgp.org.au (checked 19 July 2026): the AKT is 150 single-best-answer questions in an Australian general-practice context; the KFP is 70 independent multiple-selection questions, each case standing alone, rewarding specific and safe decisions; both are currently paper-based with computer-readable answer sheets and run around four hours each — verify your diet's delivery and timing. The CCE is the separate live consultation exam and is out of scope for a written-materials workflow. Keep official requirements and third-party claims separate at all times: only the RACGP defines the blueprint and the standard.
Step one: inventory and label every official item
List every official item you hold and tag each one: unseen (never opened), attempted once (done under some conditions), or contaminated by review (you have read the answers and now recognise the items). Only the "unseen" pile has calibration value. This inventory is the whole game — most candidates never make it, so they quietly convert their best measurement instruments into flashcards without noticing.
Step two: choose the calibration date
Pick a date late enough that a sitting is meaningful — you have covered most of the blueprint and want a true read — but early enough that you still have weeks to correct what it reveals. A calibration sitting the week before the exam tells you your result without leaving time to act on it. For most candidates a primary calibration point three to five weeks out, and a second one if genuinely unseen material remains, is the sensible rhythm.
Step three: reproduce exam conditions exactly
A calibration read is only valid if the conditions match. Sit the official sample as a single timed block, at the real per-item pace, with only the references the exam permits (for the written papers, none), no AI tutor, no pausing, breaks only where the real exam allows them, and the real response format — bubble-sheet-style answering for the AKT, and the multiple-selection discipline of the KFP where specificity is scored and scattergun answers are penalised. Anything softer inflates the number and destroys the point.
Step four: code every error by more than subject
After the sitting, code each error on three axes, not just "which topic":
- Domain — the curriculum area.
- Cognitive process — was it a knowledge gap, a data-interpretation slip, a decision-threshold error, or a misread of the task?
- Format — an AKT single-best-answer trap, or a KFP specificity failure where you were vague or over-inclusive?
The same wrong answer can mean four different things, and each points to a different fix. A knowledge gap needs learning; a threshold error needs guideline drilling; a KFP specificity failure needs practice at naming the exact, credited action.
Step five: map each error to fresh third-party practice
Every coded error becomes a prescription for daily practice — but you fill that prescription from a third-party bank, never from the official items. Keep official questions out of daily repetition entirely. Use a bank such as iatroX for the unseen volume the finite official set cannot supply, targeting the exact domain-and-process combinations your calibration exposed. The official sitting diagnoses; the bank treats.
Step six: repeat only with genuinely unseen material
Re-calibrate only when you have genuinely unseen official material — a newly released sample, a fresh examiners' report, or an official item you deliberately quarantined. If none remains, do not re-sit a contaminated set and mistake recognition for readiness; use transfer questions instead — fresh items on the same concepts, from your third-party bank, that test whether the learning generalises.
A seven-day plan: official material for calibration, iatroX for transfer
- Monday: inventory and label all official material; quarantine the unseen pile.
- Tuesday: map the blueprint against the RACGP curriculum; mark cold domains.
- Wednesday: a single official calibration sitting under exam conditions; do not peek.
- Thursday: code every error by domain, cognitive process and format.
- Friday: convert each coded error into targeted iatroX practice on fresh, unseen items.
- Saturday: a mixed, timed iatroX block to test transfer, not memory of the sample.
- Sunday: review, update your misconception log, and rest — leaving the remaining unseen official items sealed for the next calibration point.
No proprietary-algorithm claims are made here: iatroX simply supplies unseen volume and a timed transfer read, while the official material remains the scarce calibration instrument.
Decision checklist: continue, supplement, switch or stop
- Continue rationing official material if you still hold genuinely unseen items and your calibration reads are improving.
- Supplement with a third-party bank the moment daily volume demands it — which is almost immediately, because the official set is tiny.
- Switch your primary daily driver to the bank and keep official material only for calibration; never let the bank's items become your answer key to the official ones.
- Stop re-sitting contaminated official material; once recognised, an item measures memory, not readiness.
Three mistakes this workflow is designed to stop
Spending the calibration asset on revision. The commonest error is treating the official sample questions as ordinary practice — working them casually, reading the answers, and thereby converting your only clean, authoritative measure into recognisable flashcards. Once seen, they cannot be unseen. Ration them as calibration, not revision, and keep the unseen pile sealed until a calibration sitting.
Sitting official material under soft conditions. An official sample sat untimed, with notes open and no attention to pacing, produces an inflated read that tells you little. The whole value of a calibration sitting is that it matches the real four-hour, reference-free, single-block conditions; anything softer flatters you into a false sense of readiness and wastes an irreplaceable item.
Coding errors by subject alone. "I got a cardiology question wrong" is not a diagnosis. The same wrong answer can be a knowledge gap, a decision-threshold error, or a misread of the task, and each needs a different fix — learning, guideline drilling, or simply slowing down to read the lead-in. Code every error by domain, cognitive process and format, or you will keep re-treating the wrong problem.
FAQ
Is RACGP Official Resources enough for RACGP Fellowship on its own? No, and it is not designed to be. The official material is a small, high-authority set — a handful of sample questions, the curriculum, candidate guidelines and examiners' reports — that establishes the standard and the format. It cannot supply the daily question volume needed to build and test broad coverage. Its correct role is calibration and blueprint-mapping, with a third-party bank doing the repetitive practice.
Which RACGP Fellowship component does RACGP Official Resources not reproduce well? The CCE. The official written samples and curriculum documents support clinical knowledge, but they do not simulate the live, examiner-judged consultation exam. The RACGP publishes CCE candidate guidance, which you should read closely, but reading guidance is not the same as rehearsing timed, role-player consultations — that preparation happens through supervised clinical practice and dedicated CCE coaching, not through written samples.
How many RACGP Official Resources questions should I complete per day for RACGP Fellowship? Deliberately very few — ideally none on ordinary study days. Because the official item pool is so small, treating it as daily practice destroys its calibration value within a week. Ration it: reserve official items for a small number of condition-matched calibration sittings, and let a third-party bank carry your daily question volume. Scarcity is the reason to protect it, not to spend it.
When should I stop using RACGP Official Resources and move to mixed mocks? You never fully stop; you change how you use it. Throughout preparation the official material remains your calibration reference, sampled sparingly. Mixed mocks — full-length, timed, blueprint-spanning blocks from a third-party bank — should run in parallel as your daily and weekly workload, with the official set reserved for periodic true-north checks rather than volume.
How should I combine RACGP Official Resources with iatroX without duplicating practice? Assign them opposite roles. The official material is your calibration gold-standard, sat rarely and under strict conditions; iatroX is your unseen-volume and transfer-practice layer, run daily on fresh items. The one rule that prevents duplication and contamination is directional: iatroX questions may target the weaknesses the official sitting reveals, but official items must never leak into daily iatroX-style repetition, or you lose the only clean read you have.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; RACGP page contents and available samples are updated between diets — confirm current availability on racgp.org.au. Disclosure: iatroX operates a competing RACGP AKT and KFP question bank, so this article confines iatroX to the job the official materials cannot do — supplying unseen daily volume and timed transfer measurement — and treats the official set as the scarce calibration instrument it is; iatroX is not a CCE simulator. Corrections are welcome via the feedback route on iatrox.com.
References: RACGP — AKT and KFP exams, sample AKT and KFP questions, curriculum and syllabus, and candidate guidelines (racgp.org.au); iatroX — Your Q-Bank Percentage Is Not Your Exam Score; iatroX — the two-Q-bank rule; iatroX — question-bank completion is not coverage; iatroX comparison hub.