Before you tell yourself you have "covered" Internal Medicine and can stop doing new questions, you should be able to produce evidence — not a feeling. This checklist sets out the minimum evidence: a blueprint-coverage matrix filled in, ten common blind spots specifically checked, interpretation and recency audits done, and a set of performance signals met on unseen, timed material. If you cannot tick these, you have not covered the blueprint; you have covered the questions you happened to choose. This is a hub for one decision — stop or continue — and platform-specific articles should link up to it rather than repeat it.
One boundary up front. The RCPSC certification in Internal Medicine has two parts: a computer-based written multiple-choice component and a separate Applied examination in an oral or OSCE-style format, sat at the same test centre. A question bank — iatroX included — is a tool for the written-knowledge layer only. It measures and builds the knowledge the written component tests; it does not reproduce the Applied examination. Keep that split in mind throughout, because several items below apply only to the written component, and the Applied exam needs its own, different preparation.
The minimum evidence checklist
You can reasonably claim written-component coverage when all of the following are true. Treat it as a gate, not a timetable.
- Every blueprint domain has a meaningful number of attempted questions, not just the ones you enjoy.
- First-attempt accuracy on unseen items — not on review — is at or above your target in every domain, with no domain quietly left blank.
- You have sat recent, timed, mixed blocks that interleave domains, and your accuracy held up.
- Your high-confidence errors are near zero: the questions you were sure of and got wrong are the ones that fail candidates.
- Guidance-sensitive topics have been checked against current Canadian sources within the last few months.
- You have calibrated against official Royal College sample material and your in-training examination, not only commercial banks.
- Retention is demonstrated: topics learned weeks ago still return correct answers today.
Everything below is how you generate the evidence for those ticks.
RCPSC Internal Medicine in 2026: the exam snapshot
Certification has two components: a written multiple-choice examination and a separate Applied examination (oral/OSCE-style). The Royal College does not publish granular question counts, session timings or a percentage blueprint on a readily accessible page, so verify the current structure and question count on royalcollege.ca rather than relying on any third-party number, including this one. What the College does publish, and what you should build your preparation around, is the Objectives of Training in Internal Medicine and the associated competency requirements; those objectives, together with your programme's in-training examination (which is designed to mirror the certifying exam), are the authoritative map. Where a commercial bank's blueprint and the Royal College objectives disagree, the objectives win.
Both components are typically administered at the same test centre, but they assess different things — the written exam samples breadth of knowledge, the Applied exam samples reasoning under questioning — so evidence that you are ready for one is not evidence that you are ready for the other. This checklist builds the evidence for the written component; treat the Applied examination as a parallel track with its own preparation, and do not let a strong written trajectory persuade you that the oral is handled.
Because there is no public percentage blueprint, the coverage table below uses relative emphasis derived from the Objectives of Training, not official weightings. Label your own version the same way, and do not present estimated weights as if the College had published them.
Build a blueprint coverage table
Rebuild this table for yourself and fill every cell. The discipline is in the empty cells: they are exactly the domains self-selected practice skips. The data below is illustrative — invented to show the format — not a target.
| Domain (from Objectives of Training) | Relative emphasis (illustrative) | Questions attempted | First-attempt accuracy | Last reviewed | Confidence (1–5) |
|---|---|---|---|---|---|
| Cardiology | High | 180 | 74% | 12 Jul | 4 |
| Respirology | High | 140 | 70% | 10 Jul | 3 |
| Gastroenterology & hepatology | High | 130 | 68% | 8 Jul | 3 |
| Nephrology & acid–base | High | 120 | 61% | 5 Jul | 2 |
| Endocrinology | Medium | 110 | 72% | 9 Jul | 4 |
| Infectious diseases | High | 150 | 69% | 11 Jul | 3 |
| Rheumatology | Medium | 90 | 66% | 3 Jul | 3 |
| Haematology & transfusion | Medium | 85 | 58% | 1 Jul | 2 |
| Medical oncology | Medium | 70 | 63% | 2 Jul | 2 |
| Neurology | Medium | 95 | 60% | 6 Jul | 2 |
| Geriatrics & frailty | Medium | 40 | 55% | 20 Jun | 2 |
| General & ambulatory IM | High | 60 | 64% | 28 Jun | 2 |
| Ethics, capacity & goals of care | Medium | 25 | 59% | 15 Jun | 2 |
| Statistics & critical appraisal | Medium | 30 | 57% | 18 Jun | 2 |
Read the table for pattern, not just percentages. A low attempted-count with a flattering accuracy is not strength; it is an untested domain. A high count with a recent review date and mid-range accuracy that is climbing is genuine progress. The domains with both low counts and low confidence — here, geriatrics, ambulatory IM, ethics and statistics — are where you keep doing new questions.
Ten domain-level blind spots self-selected practice hides
These are the areas candidates most often leave hidden because they are less represented in popular banks or feel less like "real medicine". Check each deliberately, and have an exam-experienced internist review your coverage of them before you decide you are done.
- Ambulatory and outpatient internal medicine — most banks over-represent inpatient acute care; the exam does not.
- Geriatric medicine, frailty and deprescribing — assessment, falls, polypharmacy and stopping drugs, not just starting them.
- Preventive care and screening — Canadian screening intervals and immunisation schedules, which differ from US ones.
- Medical ethics, capacity and goals of care — consent, substitute decision-making and end-of-life planning in the Canadian legal context.
- Biostatistics and critical appraisal — likelihood ratios, number needed to treat, and reading a trial, which reliably appear and are reliably skipped.
- Dermatology for the internist — drug eruptions and the rashes that signal systemic disease.
- Allergy and clinical immunology — anaphylaxis, drug allergy and immunodeficiency.
- Palliative and end-of-life symptom management — opioid rotation, dyspnoea and delirium at the end of life.
- Addiction medicine and substance use — including the Canadian harm-reduction context.
- Transfusion medicine and anticoagulation reversal — product selection, reactions and the reversal agents for modern anticoagulants.
Format checklist: Canadian context, applied cases and oral reasoning
- Have you practised in a Canadian practice context — drug availability, provincial variation, referral pathways — rather than defaulting to a US bank's assumptions? Many strong banks are US-calibrated; verify Canadian localisation before you trust a management answer.
- Have you practised applied, multi-step cases that ask for the next best action given evolving information, not just single-fact recall?
- Have you done deliberate oral and clinical-reasoning practice for the Applied examination — spoken, structured, out loud with a colleague or supervisor? A question bank cannot rehearse this, and it is a separate exam; do not let strong written performance lull you into neglecting it.
Interpretation checklist
Internal Medicine is heavy on data interpretation, and these are testable skills that recall-only revision neglects. Confirm deliberate, timed practice on each that applies:
- ECGs — arrhythmias, ischaemia, blocks, electrolyte and drug effects.
- Chest radiographs and cross-sectional imaging — the common patterns an internist must read unaided.
- Peripheral blood films and bone-marrow clues — the morphology that points to a diagnosis.
- Arterial blood gases and acid–base — full mixed-disorder problems, not just the easy ones.
- Laboratory trends — renal, hepatic and electrolyte trajectories over time, not isolated values.
- Calculations — anion gap, corrected calcium, A–a gradient, creatinine clearance, and osmolar gap.
- Statistics and ethics vignettes — interpreting a study and reasoning through a capacity or consent scenario.
Recency checklist: the guidance-sensitive topics
Some answers change with the guidelines, and a bank written two years ago may be wrong now. For each guidance-sensitive topic, record the date and jurisdiction of the source you are trusting, and prefer current Canadian guidance where it exists (for example, Diabetes Canada, the Canadian Cardiovascular Society, Hypertension Canada, Thrombosis Canada). Check at minimum:
- heart-failure therapy and the current pillars of treatment;
- diabetes management, including agents with cardiorenal indications;
- anticoagulation and the reversal of direct oral anticoagulants;
- sepsis and antimicrobial stewardship;
- lipid and hypertension targets;
- chronic kidney disease management;
- immunisation and infection-prevention guidance.
If you cannot name the source and its year for one of these, that topic is not covered — it is remembered, which is different.
Performance checklist: the signals that actually predict readiness
Coverage is necessary but not sufficient; readiness is a performance property. Confirm all of these on unseen material:
- Unseen, timed, mixed blocks — your accuracy holds when domains are interleaved and the clock is running, not just in single-topic review mode.
- Speed — you are finishing within the per-question pace the written component demands, without a late-paper collapse.
- High-confidence errors near zero — you have hunted down the items you were sure of and got wrong, because those are the dangerous ones.
- Retention — material from weeks ago still returns correct answers on re-test, not just on the day you learned it.
- Official-material calibration — you have measured yourself against Royal College sample questions and your in-training examination, and the result agrees with your bank performance. If they disagree, trust the official material.
Your overall q-bank percentage is not among these signals, and it is not your exam score. It blends seen and unseen items, easy and hard domains, and reviewed and first-attempt questions into a single number that flatters you. Measure the components above instead (the linked explainer makes the case in full).
Stop or continue? A decision tree
Use the measured gap, not the calendar or your fatigue, to choose the next action:
- Continue new questions if any blueprint domain is thin on attempts or below target on first-attempt unseen accuracy. Unseen coverage is still incomplete.
- Consolidate (stop new questions, review errors, space the misses) if coverage is broad and accuracy is adequate but retention is shaky or high-confidence errors persist. More new questions will not fix a retention problem.
- Simulate if written performance is solid but the Applied examination is under-rehearsed — shift to oral and clinical-reasoning practice with colleagues.
- Seek teaching if one domain resists improvement despite volume; a targeted session with an expert beats another hundred questions.
- Rest if accuracy is high, retention is proven and errors are careless rather than knowledge-based — you are ready, and fatigue is now the main threat to your score.
The one-page checklist and a worked example
Copy this into your notes and fill it in; it is the whole method on one page.
- Every blueprint domain has an attempted-question count above your threshold.
- First-attempt unseen accuracy meets target in every domain — no blanks.
- All ten common blind spots specifically checked and reviewed by an experienced internist.
- Interpretation skills (ECG, imaging, blood film, acid–base, labs, calculations) practised under time.
- Guidance-sensitive topics dated and sourced to current Canadian guidance.
- Recent timed mixed blocks sat; accuracy held under interleaving.
- High-confidence errors reviewed to near zero.
- Calibrated against Royal College sample material and the in-training exam.
- Applied-examination practice under way separately.
Worked example, with invented data. A candidate has completed about 1,400 questions at a 71% overall bank average and feels ready to stop. The matrix tells a different story: nephrology sits at 61% first-attempt unseen, geriatrics has only 40 attempts, and ethics and statistics are both under 60% with fewer than 30 attempts each. Two recent timed mixed blocks came in at 64%, below the single-topic average, and three high-confidence errors recurred in anticoagulation reversal. The decision is not "stop" but "continue, narrowly": new questions in nephrology, geriatrics, ethics and statistics; a dated review of anticoagulation guidance; and a second look at why the mixed-block score trails the topic-by-topic one. The overall 71% was hiding all of this — which is exactly what an overall percentage does. Note too that the ambulatory and geriatrics gaps would not have surfaced without the matrix, because a self-directed learner rarely chooses to drill the domains they find least engaging; the matrix makes the avoidance visible.
Frequently asked questions
How do I know whether I have covered the full RCPSC Internal Medicine blueprint? You know when a completed blueprint-coverage matrix — built from the Royal College Objectives of Training, not a commercial bank's table — shows a meaningful attempted-question count and an at-target first-attempt unseen accuracy in every domain, with the ten common blind spots specifically checked and an experienced internist having reviewed your coverage of the less popular areas. Coverage is a property of the map, not of any one bank; if whole domains are thin or blank, you have covered your favourite questions rather than the blueprint, however high your overall percentage looks.
Can one question bank be enough for RCPSC Internal Medicine? One bank can be enough for breadth of the written component only if you have verified that its blueprint genuinely matches the Royal College objectives and that its content is Canadian-calibrated rather than quietly US-oriented, and only if you calibrate against official Royal College sample material and your in-training exam rather than trusting the bank's own accuracy figure. Even then, a single bank becomes recognition once you have cycled it, which is why a second source of unseen questions is usually worth it — and no bank covers the separate Applied examination, which needs its own oral and clinical-reasoning practice.
What should I measure instead of my overall Q-bank percentage for RCPSC Internal Medicine? Measure first-attempt accuracy on unseen items broken down by blueprint domain, your accuracy on recent timed mixed blocks, your per-question pace, your count of high-confidence errors, and your retention on delayed re-tests — then check all of that against official Royal College sample material. The overall percentage blends seen and unseen, easy and hard, and reviewed and first-attempt questions into one flattering figure; the component signals are what actually track readiness, and they will often disagree with the headline number in ways that change your plan.
When should I stop doing new RCPSC Internal Medicine questions? Stop starting new questions when every domain meets its coverage and accuracy target on unseen material, your timed mixed blocks hold up, your high-confidence errors are near zero and your retention is proven — at which point more new questions have a low marginal return and your time is better spent consolidating errors, rehearsing the Applied examination, or resting. Until those conditions are met, a thin or below-target domain is a signal to keep going, regardless of how many questions you have already done or how tired you are.
Which RCPSC Internal Medicine resource should I use for my weakest component? Match the resource to the component: for a weak written-knowledge domain, use unseen, Canadian-calibrated questions and a current guideline or standard reference for that topic; for weak data interpretation, use deliberate ECG, imaging and acid–base practice under time; and for the Applied examination, use spoken, structured oral and clinical-reasoning practice with a colleague or supervisor, because no question bank — iatroX included — reproduces the oral format. iatroX fits the written-knowledge and unseen-measurement job specifically, so use it there and reach for the oral-practice and interpretation tools where they are the right fit.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Exact question counts, session timings and blueprint weightings for the RCPSC Internal Medicine examination are not published on a readily accessible page; the emphasis figures in this article are illustrative estimates drawn from the Objectives of Training, not official weightings, and should be verified against royalcollege.ca. Any vendor figures mentioned are vendor-reported and were current at the last-checked date. Disclosure: iatroX operates a competing question bank and clinical-knowledge platform; this article confines iatroX's role to the written-knowledge and unseen-measurement layer, and iatroX does not reproduce the separate Applied examination. Corrections are welcome via the feedback route on iatrox.com.
References: Royal College of Physicians and Surgeons of Canada — Internal Medicine examination format and Objectives of Training (royalcollege.ca); iatroX Canada exam hub (iatrox.com/canada); "Your Q-Bank Percentage Is Not Your Exam Score"; the completion-is-not-coverage and two-Q-bank rule framework pillars; the iatroX comparison hub.
Complete a fresh RCPSC Internal Medicine baseline in iatroX →