This is the checklist to work through before you tell yourself you have "covered" the AMC CAT MCQ and can stop doing new questions. It is written for international medical graduates on the standard pathway who have already done a large volume of practice and now need to decide, honestly, whether the volume has produced coverage or just comfort. The principal risk is not laziness; it is a high overall percentage sitting on top of two or three untested domains.
Coverage is a claim you should have to prove. The rest of this article is the evidence you need before you can make it.
The direct answer: what "covered" actually requires
You have covered the AMC MCQ when, and only when, you can show all six of the following — not a study timetable, but a set of pass conditions:
- Every blueprint domain has been attempted in enough volume to give a stable first-attempt accuracy, not a lucky handful of items.
- No domain sits below your personal safety threshold on unseen, first-attempt questions.
- Your accuracy holds under timed, mixed conditions — roughly 84 seconds per item — not only in untimed, single-topic sets.
- Your high-confidence errors (answers you were sure of and got wrong) are near zero, because those are the ones that sink a score.
- Guidance-sensitive topics have been checked against current Australian sources, with the date and jurisdiction recorded.
- Your performance has been calibrated against the AMC's own material, not only a commercial bank.
If you cannot yet show all six, you have not covered the exam — you have practised part of it. The sections below turn each into something you can tick.
Current exam snapshot
The AMC CAT MCQ is 150 single-best-answer questions, one correct response from five options, delivered in one 3.5-hour session, computer-administered at Pearson VUE test centres. It is named a "Computer Adaptive Test (CAT)," but the AMC's published specification does not confirm that item difficulty adapts to your performance, so plan for a fixed-length paper of 150 items and a steady pace of about 84 seconds per question — do not assume the test will shorten, personalise or ease off if you do well. All 150 items count toward your result under the published specification; confirm on the AMC page whether any trial items are included in the sitting you book.
The content spans essential medical knowledge across the disease process, clinical examination and diagnosis, investigation, and therapy and management, drawn from the full breadth of medicine, surgery, paediatrics, obstetrics and gynaecology, psychiatry and population health. From 2026 the reported pass mark is 250 on a 0–500 scale, with a slightly higher cut score introduced that year; the standard is pitched at the level of a graduating Australian medical student. The MCQ is the gateway to the AMC Clinical Examination: a pass certifies readiness for the next stage rather than for independent practice, which is one reason the blueprint is broad and shallow rather than deep in any single discipline — a shape your revision has to mirror rather than fight. The authoritative sources are the AMC MCQ specifications and the official candidate handbook on amc.org.au — treat everything else, including this article, as commentary on those documents.
Build a blueprint coverage table
The single most useful artefact you can build is a coverage table with one row per domain and these columns: official emphasis, questions attempted, first-attempt accuracy, date last reviewed, and confidence. It converts a vague sense of readiness into something auditable. Here is the structure, with clearly invented illustrative data for one candidate, "Dr A," to show how it reads.
| Domain | Official emphasis | Qs attempted | First-attempt accuracy | Last reviewed | Confidence |
|---|---|---|---|---|---|
| General medicine | High | 420 | 74% | 6 days ago | Medium |
| General surgery | High | 260 | 71% | 9 days ago | Medium |
| Paediatrics | Moderate | 90 | 58% | 22 days ago | Low |
| Obstetrics & gynaecology | Moderate | 85 | 55% | 24 days ago | Low |
| Psychiatry | Moderate | 70 | 68% | 12 days ago | Medium |
| Population & preventive health | Moderate | 30 | 61% | 30 days ago | Low |
| Ethics, law & professionalism | Threaded | 25 | 64% | 28 days ago | Low |
Dr A's overall percentage would look reassuring, because the two high-volume medical and surgical domains dominate the average. The table exposes the truth: paediatrics, obstetrics and gynaecology, population health and ethics are under-sampled, stale and weak. A completion figure would hide all of that; a coverage matrix makes it impossible to ignore. If you want the full method for building one for any exam, our pillar on why completion is not coverage sets it out.
The ten domain-level blind spots most likely to stay hidden
Self-selected practice hides gaps in exactly the areas you find least comfortable. For the AMC MCQ, these ten are the most common blind spots, and each should be checked by someone with exam-specific clinical knowledge before you conclude you are safe — do not sign off your own weak areas.
- Antenatal and intrapartum care — the normal antenatal screening schedule, common obstetric emergencies and their first management step.
- Child health — the Australian immunisation schedule, developmental milestones, and weight-based paediatric prescribing.
- Aboriginal and Torres Strait Islander health — culturally safe care, the chronic-disease burden, and specific screening considerations.
- Mental health and the law — risk assessment, capacity, and involuntary treatment under state Mental Health Act provisions.
- Sexual health and notifiable diseases — STI management, contact tracing and Australian notification requirements.
- Skin cancer — melanoma, squamous and basal cell carcinoma, which are disproportionately important in the Australian context.
- Prescribing in older adults — polypharmacy, interactions, and deprescribing.
- Acute and emergency recognition — sepsis, acute coronary syndromes and the "next best step" under time pressure.
- Ethics and professionalism — consent, mandatory reporting, open disclosure and confidentiality within Australian frameworks.
- Population and preventive health — the national bowel, breast and cervical screening programmes and structured chronic-disease management.
Format checklist
Coverage of content is not enough; you must also have practised the exam's specific demands. Verify that you have deliberately rehearsed:
- Australian guideline localisation. Your default answers should reflect Australian practice — Therapeutic Guidelines (eTG) and the Australian Medicines Handbook for therapy, national screening programmes for prevention — not the guidelines of the country you trained in. This is the single most common source of confident, wrong answers for IMGs.
- The IMG transition. Watch for topics where your prior training diverges from Australian norms: first-line drug choices, referral pathways, thresholds for investigation, and terminology.
- Fixed-length pacing. Because the paper is 150 items in 3.5 hours and you should not assume it adapts, rehearse a metronomic pace of roughly 84 seconds per item, including flagging and returning, so time management is automatic.
Interpretation checklist
Work through whether you have practised, on unseen items, each interpretive task the exam can set: clinical photographs and rashes; ECGs; plain radiographs and common CT findings; laboratory trends over time rather than a single value; simple clinical calculations; and any items that turn on ethics or basic statistics. A domain can be "covered" in text yet fail you the moment it is presented as an image or a trend, so test the format, not just the fact.
Recency checklist
Some topics move. Identify every guidance-sensitive area — screening intervals, immunisation updates, antimicrobial recommendations, chronic-disease targets — and for each, record the date and jurisdiction of the source you last learnt it from. An answer that was correct two years ago or in another country is a liability. If you cannot name a current Australian source and its date, mark the topic unverified and re-check it.
Performance checklist
Finally, verify the qualities of your performance, not just its coverage:
- Unseen, timed, mixed blocks. Your headline evidence should come from blocks of questions you have never seen, sat under time, with domains interleaved — because that is the exam.
- Speed. Accuracy that only appears when untimed will not survive 84 seconds per item.
- High-confidence errors. Track these separately; they are the most dangerous category and the best predictor of an unpleasant surprise.
- Retention. Re-test old topics after a gap to confirm the knowledge is durable, not freshly crammed.
- Official-material calibration. Sit the AMC's own practice materials unseen and timed at least once, as the closest available proxy for the real standard.
Stop or continue: a decision tree
When you have the evidence above, the next action follows from the measured gap, not from habit:
- Continue new questions only where a domain is under-sampled or below your threshold on unseen items. New volume is for genuine gaps, not for the comfort of a rising counter.
- Consolidate — stop new questions and review — where accuracy is adequate but high-confidence errors or poor retention show the knowledge is unstable.
- Simulate — full-length, mixed, timed papers — where individual domains are solid but you have never tested stamina and pacing across 150 items.
- Seek teaching where a domain resists self-study, particularly the localisation gaps an IMG cannot easily self-correct.
- Rest where every domain is covered, calibrated and stable; more questions past that point buy fatigue, not marks.
A one-page checklist to copy
Copy this and fill it in; when every box is ticked with evidence, you have earned the right to stop doing new questions.
- Every domain attempted in stable volume, logged in a coverage table
- No domain below threshold on unseen, first-attempt items
- Accuracy holds under timed, mixed, fixed-length conditions (~84s/item)
- High-confidence errors tracked and near zero
- Ten common blind spots each checked by someone with exam knowledge
- Australian localisation and IMG-transition traps rehearsed
- Image, ECG, radiograph, lab-trend and calculation formats practised unseen
- Guidance-sensitive topics dated and sourced to current Australian references
- AMC official material sat unseen and timed for calibration
A worked example (invented data). Dr A, from the table above, has an overall accuracy near 70% and feels ready. Working the checklist, she finds four domains stale and below threshold, no image practice logged for dermatology, and antenatal screening last learnt from her original training abroad. Her decision is not "keep doing new questions across the board." It is: consolidate general medicine and surgery (stop new questions there), do targeted new questions in the four weak domains, add unseen image sets, re-source her obstetric knowledge to Australian guidance, and sit one official practice paper timed. The checklist turned a vague "nearly ready" into five specific, bounded actions.
Two weeks later, Dr A re-runs the table. Paediatrics has moved from 58% to 71% on unseen items and obstetrics from 55% to 69%, but her high-confidence error log has surfaced a new pattern: she keeps confidently choosing overseas first-line antihypertensives and antibiotics. That is not a knowledge-volume problem, so more questions would not fix it; the checklist redirects her instead to a focused localisation review against Australian guidance and a re-test of only those items. Her population-health row, still at 61% and last reviewed a month earlier, trips the recency check rather than new practice. The value of the exercise is not that it produced a pass prediction — it did not, and no honest checklist should — but that at every step it named the specific cause and matched it to the single activity that addresses it.
Frequently asked questions
How do I know whether I have covered the full AMC MCQ blueprint? You know only when a blueprint coverage table shows every domain attempted in stable volume, each meeting your accuracy threshold on unseen and timed items, with no stale or unverified rows. A high overall percentage does not demonstrate coverage, because it is dominated by whichever domains you happened to practise most; coverage is a per-domain property, and it has to be shown row by row against the official emphasis, not inferred from an average.
Can one question bank be enough for AMC MCQ? One well-built bank can carry the bulk of your learning, but a single bank cannot tell you whether your knowledge transfers, because by the end you have seen its items and are partly recognising them. The safest structure is one bank for volume and a second, unseen source used purely for measurement, plus the AMC's own material for calibration — the point is not more questions but an independent check that your score is real rather than a memory of that bank.
What should I measure instead of my overall Q-bank percentage for AMC MCQ? Measure per-domain first-attempt accuracy on unseen items, your high-confidence error rate, and your accuracy under timed mixed conditions. The overall percentage is a blended, backward-looking number that hides weak domains and rewards repetition; the reasons it misleads are set out in Your Q-Bank Percentage Is Not Your Exam Score. Track the components that predict exam-day behaviour instead.
When should I stop doing new AMC MCQ questions? Stop adding new questions in a domain once it is covered, calibrated and stable on unseen timed items, and redirect that time to consolidation, full-length simulation or rest. Continuing to grind new questions across already-solid domains is the most common way strong candidates waste their final weeks; new volume should be reserved for domains the coverage table still flags as under-sampled or weak.
Which AMC MCQ resource should I use for my weakest component? Match the resource to the deficit. If the weakness is Australian localisation, go to the primary Australian guidance (Therapeutic Guidelines, national screening programmes) rather than more questions; if it is a knowledge gap, use focused unseen questions in that domain; if it is pacing, use timed full-length simulation; and if it is a specific interpretive skill such as ECGs or radiographs, use format-specific unseen sets. The right resource is the one that targets the measured cause, not the most popular bank.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; exam figures reflect the AMC's published specification on that date — the pass mark, cut score and any inclusion of trial items can change, so verify the current specification and handbook on amc.org.au. This article deliberately describes the AMC CAT MCQ as a fixed-length computer-administered paper, because the AMC page does not confirm that difficulty adapts to performance despite the "CAT" name; do not plan around adaptivity. Disclosure: iatroX operates a competing AMC MCQ question bank; its role here is confined to being the unseen-measurement and syllabus-audit layer this checklist relies on, and its overall percentage is subject to exactly the same caveat as any other bank's. Corrections are welcome via the feedback route on iatrox.com.
References: AMC MCQ specifications and candidate handbook (amc.org.au); Pearson VUE AMC information (pearsonvue.com); iatroX internal references — Your Q-Bank Percentage Is Not Your Exam Score, question-bank completion is not coverage and the AMC MCQ bank landing page.