Before you say you have "covered" RACGP Fellowship, you need evidence, not a finished progress bar. This checklist sets out the minimum proof required across the three components — the Applied Knowledge Test, the Key Feature Problem and the Clinical Competency Exam — so you can decide, from measured data, whether to keep doing new questions or move on. It is deliberately a checklist rather than a study timetable: the point is what you can demonstrate, not how many weeks you have spent. Run a fresh baseline in iatroX and use the decision tree at the end to choose the next activity.
The minimum evidence to claim coverage
You can reasonably say you have covered RACGP Fellowship when you can show all of the following, and not before: a blueprint coverage table with real accuracy data in every domain; deliberate practice in each of the three exam formats, not just the AKT; evidence you can interpret images, data and calculations under time; a recency check on the guidance-sensitive topics; and performance data from unseen, timed, mixed blocks rather than untimed single-topic practice. Each is a line you tick with evidence, and the rest of this article is how to gather it.
The exam you are actually preparing for
RACGP Fellowship (FRACGP) has three assessments. The Applied Knowledge Test is 150 single-best-answer questions. The Key Feature Problem is 70 multiple-selection questions, each case independent of the others. Both are currently delivered on paper with computer-readable answer sheets; confirm the current delivery mode and durations on racgp.org.au, because these are reviewed periodically. The Clinical Competency Exam is the clinical, consultation-based assessment; the College sets its structure, and you should read the current CCE handbook rather than assume a fixed format from any third party. These figures were last checked on 19 July 2026 against RACGP material.
The authoritative sources are the RACGP curriculum, the examination handbooks and the College's official practice material. A question bank is a training tool built against that curriculum; it is not the curriculum, and — importantly — no bank is a CCE simulator. iatroX covers the AKT and KFP-style knowledge and reasoning; it is a written-knowledge and unseen-measurement layer, not a substitute for supervised clinical assessment.
Build a blueprint coverage table
The core artefact of this checklist is a coverage table you fill with your own data. One row per curriculum domain, five columns.
| Domain (example) | Official weight | Questions attempted | First-attempt accuracy | Last reviewed | Confidence (H/M/L) |
|---|---|---|---|---|---|
| Cardiovascular | — | 60 | 72% | 12 Jul | M |
| Mental health | — | 25 | 55% | 2 Jun | L |
| Women's health / antenatal | — | 18 | 61% | 20 Jun | L |
| Paediatrics & child health | — | 30 | 68% | 8 Jul | M |
| Aboriginal & Torres Strait Islander health | — | 8 | — | never | L |
| Aged care & multimorbidity | — | 22 | 64% | 1 Jul | M |
| Dermatology | — | 15 | 70% | 5 Jul | M |
| Musculoskeletal | — | 20 | 66% | 3 Jul | M |
The figures above are illustrative. The discipline is real: a domain with few attempts, an old review date and low confidence is an unmeasured gap regardless of your overall percentage. Fill the weight column from the current RACGP curriculum rather than guessing.
Ten domain-level blind spots self-selected practice tends to hide
Left to choose your own questions, you will drift toward what you already know. These ten areas are the ones most often left thin, and they warrant exam-specific clinician review before you conclude you have covered them: Aboriginal and Torres Strait Islander health; rural and remote presentations; mental health and risk assessment; women's health and antenatal shared care; sexual health and contraception; child development and safeguarding; aged care, frailty and polypharmacy; palliative and end-of-life care in the community; ethical, legal and medico-legal scenarios specific to Australian general practice; and multimorbidity where two guidelines conflict. A domain being uncomfortable is exactly why it hides.
Format checklist: practise all three, not just the AKT
Single-best-answer volume is the easy habit; it is not the whole exam. Verify deliberate practice in each format. Key-feature reasoning: the KFP rewards identifying the small number of decisions that actually change the outcome of a case, and it is a distinct skill from AKT recognition — practise it as its own thing. Australian primary-care guidance: confirm your answers reflect current Australian sources and RACGP position, not imported overseas thresholds. Consultation skills: the CCE assesses real consulting — data gathering, management, communication and safety in a time-limited encounter — which is trained through observed practice and video review, not through more MCQs.
Interpretation checklist
Tick each of these against timed practice, because they fail quietly under exam pressure: image interpretation (rashes, fundi, clinical photographs); ECGs; radiographs and imaging reports; laboratory trends over time rather than a single value; clinical calculations done accurately at speed; and the ethics and basic statistics or evidence-appraisal content the blueprint includes. Each of these is a skill that degrades under a ticking clock even when the underlying knowledge is sound, which is exactly why untimed practice flatters it. If you have never done these under time, you have not tested them — you have only confirmed that you could do them with time to spare, which is not the condition the exam imposes.
Recency checklist
Guidance moves, and a bank can lag. Identify the guidance-sensitive topics — cardiovascular risk, diabetes, contraception, antimicrobial choice, mental health prescribing, immunisation schedules — and for each record the date and jurisdiction of the source you are relying on. A correct-but-superseded answer is a real risk in primary care, where the standard is current Australian practice. If your source is old or overseas, that line is not yet ticked.
Performance checklist
Finally, verify the performance evidence, because knowledge without performance does not pass. You want data on: unseen, timed, mixed blocks at exam length, not single-topic untimed sets; speed, particularly your pace on the KFP; your high-confidence error rate, because confident mistakes are the ones you will not catch yourself; retention, measured by re-testing old material cold weeks later; and calibration against the RACGP official practice material, which is the closest thing to the real standard. When these are in hand, your readiness signal is a rising unseen score, not a full progress bar.
Stop or continue: a decision tree
Use the measured gap to choose one action, not several.
| If your data shows… | Then… |
|---|---|
| Domains with few attempts or low accuracy | Continue new questions, targeted at those domains |
| Broad coverage but errors are pace or misreading | Consolidate: timed mixed blocks, not more new topics |
| Knowledge solid but consultation skills untested | Simulate: observed CCE-style practice and video review |
| A concept you cannot self-correct | Seek teaching on that specific concept |
| Rising fatigue and falling accuracy near the diet | Rest and light review; protect the exam-day performance |
The wrong move is to keep grinding new AKT questions because it is the comfortable habit when your real gap is the KFP, the CCE or your pace.
A one-page checklist and a worked example
Copy this and fill it in. Coverage: every domain has attempts, accuracy and a recent review date. Format: AKT single-best-answer, KFP key-feature reasoning and CCE consultation practice all done. Interpretation: images, ECGs, radiographs, lab trends, calculations tested under time. Recency: guidance-sensitive topics dated and jurisdiction-checked. Performance: unseen timed mixed blocks, KFP pace, high-confidence errors, retention and official-material calibration all measured.
Worked example, using invented data. A registrar reports an overall bank score of 74% and feels ready. The table tells a different story: mental health 55% with a June review, women's health 61% with only 18 attempts, and Aboriginal and Torres Strait Islander health never reviewed. KFP practice is thin, and no CCE-style consultation has been observed in a month. The honest read is not "74%, nearly there"; it is three unmeasured domains, one under-practised format and an untested clinical component. The next actions choose themselves: targeted questions in the three weak domains, dedicated KFP reasoning practice, and a booked observed consultation — not another mixed AKT block.
Follow that registrar forward two weeks and the value of reading the table rather than the headline becomes concrete. She spends the first week on the three weak domains until each has at least thirty timed attempts and a fresh review date, watching first-attempt accuracy rather than her comfort. She spends the KFP sessions specifically hunting the key decisions in each case instead of answering it as though it were an AKT item. She books two observed consultations and reviews the recordings against the CCE domains. At the end of the fortnight her overall percentage has barely moved — but her coverage table is no longer hiding three domains, her KFP pace has a baseline, and her consulting has been seen by another clinician. The number that mattered was never the 74%; it was the pattern underneath it.
Three mistakes this checklist is designed to stop
The first mistake is treating bank completion as coverage. Finishing every question in a bank tells you that you have seen the bank, not that you have covered the RACGP curriculum; the domains you rushed and the ones the bank happens to under-represent are invisible in a completion percentage. The coverage table exists precisely to surface those gaps.
The second mistake is measuring readiness on the AKT alone because it is the most bank-friendly component. A candidate who is strong on single-best-answer recognition can still be under-prepared for the KFP's key-feature reasoning and entirely untested on the CCE's consulting, and averaging all three into one confident number hides both weaknesses. Score the components separately or do not score them at all.
The third mistake is trusting an old or overseas answer as current Australian practice. Primary care is guidance-sensitive, and a bank explanation written two guideline cycles ago, or against a United Kingdom or United States threshold, can be confidently wrong for RACGP. The recency column is not bureaucracy; it is the line between a defensible answer and a superseded one.
Each of these mistakes shares a root: substituting a comfortable proxy — completion, an overall percentage, a remembered fact — for measured, current, per-component evidence. The checklist is the antidote.
Bottom line
The honest test of readiness for RACGP Fellowship is not how many questions you have finished but what you can demonstrate: measured first-attempt accuracy in every domain, deliberate practice in all three formats, interpretation and calculation tested under time, guidance-sensitive topics dated against current Australian sources, and a rising unseen score calibrated to the College's own practice material. When those lines are ticked, stop adding volume and consolidate. When they are not, the empty lines tell you exactly what to do next — targeted questions, KFP reasoning, observed consulting, teaching or rest — which is the whole purpose of auditing coverage rather than trusting a percentage.
Frequently asked questions
How do I know whether I have covered the full RACGP Fellowship blueprint? You know when your coverage table shows attempts, first-attempt accuracy and a recent review date in every curriculum domain, when you have practised all three formats rather than only the AKT, and when your performance data comes from unseen timed blocks calibrated against the official RACGP practice material. Coverage is what your evidence demonstrates across the blueprint, not what a bank's completion percentage claims, and the domains you have avoided are precisely the ones a progress bar hides.
Can one question bank be enough for RACGP Fellowship? One strong bank can cover most of the AKT and KFP knowledge work, but no single bank is enough by itself. You should not measure your readiness on the same items you learned from, so an independent unseen set is needed for honest calibration; and the CCE is a consultation exam that no bank simulates, so observed clinical practice is non-negotiable. Treat a bank as the core of the written preparation, with unseen measurement and real consulting practice added around it.
What should I measure instead of my overall Q-bank percentage for RACGP Fellowship? Measure first-attempt accuracy per domain on unseen, timed blocks; your pace, especially on the KFP; your high-confidence error rate; your retention on re-tested old material; and your accuracy against the RACGP official practice questions specifically. Your blended overall percentage — inflated by re-seen items and untimed practice — is the least useful figure you hold, as the standard caveat article explains; the per-domain unseen trend is what actually predicts readiness.
When should I stop doing new RACGP Fellowship questions? Stop adding new questions when new ones stop revealing new gaps — when your errors are pace, misreading or format-specific rather than missing knowledge — and switch to consolidation, KFP-specific reasoning and CCE-style consultation practice. Grinding more AKT items past that point is the comfortable option, not the effective one; the marginal gain has moved to the formats and skills you have been avoiding.
Which RACGP Fellowship resource should I use for my weakest component? Match the resource to the deficit. For AKT knowledge breadth, use a large single-best-answer bank with an independent unseen set for measurement. For KFP reasoning, use key-feature-format practice, because it is a distinct skill from recognition. For the CCE, use observed consultations, video review and a supervisor's feedback, since it is a real consulting assessment. Diagnose the weakest component from your coverage table and unseen data, then apply the resource built for it rather than defaulting to more MCQs.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Exam figures are RACGP-reported and reviewed periodically; confirm the current AKT and KFP question counts, delivery mode and durations, and the CCE structure, on racgp.org.au before relying on any number here. Disclosure: iatroX operates a competing question bank covering RACGP AKT and KFP-style knowledge; it is not a CCE simulator, and this checklist confines iatroX's role to the unseen-measurement and knowledge job that the clinical exam does not cover. Corrections are welcome via the feedback route on iatrox.com.
References: RACGP — Fellowship exams, AKT and KFP pages, examinations handbook and curriculum (racgp.org.au); iatroX RACGP Fellowship bank (/australia/exam/au-racgp); "Your Q-Bank Percentage Is Not Your Exam Score" (iatrox.com/blog/qbank-percentage-not-your-exam-score); completion is not coverage (iatrox.com/blog/question-bank-completion-is-not-coverage-how-to-build-a-blueprint-coverage-matrix-for-any-medical-exam); the two-Q-bank rule (iatrox.com/blog/the-two-q-bank-rule-how-to-add-a-second-bank-without-duplicating-questions-or-destroying-calibration); iatroX comparison hub (/compare).