The MRCP Part 1 Q-Bank Content-Gap Checklist: What to Verify Before You Stop Doing New Questions

This is for MRCP Part 1 candidates trying to decide whether to keep grinding new questions or stop and consolidate. The honest answer is that neither your bank percentage nor the number of questions remaining tells you. You have genuinely covered Part 1 when you can evidence six things: a representative spread across the whole blueprint, the right cognitive level, fluent data interpretation, current UK-referenced answers, deliberate work on the basic sciences most people avoid, and a stable score on fresh, timed, mixed blocks. This checklist makes each one verifiable rather than a feeling.

The six-part coverage test

Before the detail, here is the whole checklist in a single view. A question bank has done its job for MRCP Part 1 only when you can tick all six rows honestly. Everything else in this article is simply how to gather the evidence for each one.

#	What to verify	You can tick it when…
1	Blueprint coverage	You have attempted a representative sample of every specialty in the MRCP(UK) Part 1 blueprint, weighted roughly as the exam weights them — not just the topics you enjoy.
2	Cognitive level	Your practice tests application, integration and next-best-step reasoning, not only single-fact recall.
3	Data interpretation	You can read ECGs, radiographs, blood films, dynamic endocrine tests and laboratory trends at exam pace.
4	Jurisdiction and currency	Your answers reflect current UK practice, and you can name the date and source behind any guidance-sensitive item.
5	The under-practised core	You have deliberately covered clinical sciences, statistics, pharmacology and the small specialties instead of skipping them.
6	Unseen measurement	Your readiness is proven on fresh, timed, mixed blocks you have never seen — not on a re-tested bank percentage.

Most candidates can tick rows 1 and 6 by feel and are wrong about both. The rest of this article converts each row into something you can check off with data.

What MRCP Part 1 actually tests

MRCP(UK) Part 1 is delivered as two three-hour papers of 100 best-of-five (single-best-answer) questions each — 200 items in total — with no negative marking. Because wrong answers are not penalised, there is never a reason to leave a question blank, and your job in preparation is coverage and calibration rather than gambling strategy.

The examination is built on a published blueprint that fixes roughly how many items come from each area. Across the two papers the indicative distribution is: clinical sciences 25, clinical pharmacology and therapeutics 15, then cardiology, infectious diseases, neurology, renal medicine, respiratory medicine, rheumatology, gastroenterology and endocrinology at 14 each, haematology 10, psychiatry 9, dermatology 8, geriatric medicine 8, oncology 5, medical ophthalmology 4 and palliative medicine 4. Clinical sciences itself spans cell and molecular biology, clinical anatomy, biochemistry and metabolism, physiology, genetics, immunology and medical statistics — the exact material candidates most reliably neglect.

Two features of that blueprint should shape how you audit your revision. First, the eight core medical specialties at 14 items each dominate the paper, so weak coverage in any one of them is expensive. Second, the "small" areas — medical ophthalmology and palliative medicine at four items, oncology at five — are still worth marks that strong candidates collect and weaker ones concede. A bank percentage cannot tell you whether your marks are spread the way the blueprint spreads them; only a coverage matrix can.

Build a blueprint coverage matrix, not a percentage

The single most useful thing you can do before deciding to stop is to build a blueprint coverage matrix. The method is set out in full in the completion-is-not-coverage pillar, and the short version is this: list every blueprint domain as a row, then record five things against each — the official weight, the number of questions you have actually attempted, your first-attempt accuracy, the date you last reviewed the topic, and a red/amber/green confidence flag.

Domain	Official weight (items)	Questions attempted	First-attempt accuracy	Last reviewed	Confidence
Clinical sciences	25	—	—	—	R / A / G
Clinical pharmacology & therapeutics	15	—	—	—	R / A / G
Cardiology	14	—	—	—	R / A / G
Infectious diseases	14	—	—	—	R / A / G
Neurology	14	—	—	—	R / A / G
Renal medicine	14	—	—	—	R / A / G
Respiratory medicine	14	—	—	—	R / A / G
Rheumatology	14	—	—	—	R / A / G
Gastroenterology	14	—	—	—	R / A / G
Endocrinology	14	—	—	—	R / A / G
Haematology	10	—	—	—	R / A / G
Psychiatry	9	—	—	—	R / A / G
Dermatology	8	—	—	—	R / A / G
Geriatric medicine	8	—	—	—	R / A / G
Oncology	5	—	—	—	R / A / G
Medical ophthalmology	4	—	—	—	R / A / G
Palliative medicine	4	—	—	—	R / A / G

The matrix immediately exposes what an overall percentage hides. A domain can look fine because you scored well on a handful of items you happened to attempt, while the "questions attempted" column shows you have barely sampled it. Any row that is high-weight, low-attempted and amber or red is a genuine content gap; a row that is low-weight and green can be left alone. This is coverage thinking, and it is what turns "I have done 80% of the bank" into an actual readiness decision.

The ten blind spots self-selected practice hides

When candidates choose their own questions, the same areas go under-practised almost every time. Treat the following ten as high-suspicion gaps that deserve a deliberate second look — ideally reviewed against exam-specific material or by a colleague who has recently sat the paper, rather than trusted on your own read.

Clinical sciences as a whole. Twenty-five items — more than any single specialty — yet routinely the least-practised block because it feels like preclinical revision.
Medical statistics and epidemiology. Sensitivity, specificity, likelihood ratios, number needed to treat, study design and bias appear reliably and are learnable, but rarely drilled.
Genetics and inborn errors of metabolism. Inheritance patterns, trinucleotide repeats and enzyme defects sit inside clinical sciences and are easy marks once revised.
Immunology. Complement, hypersensitivity mechanisms, cytokines and immunodeficiency underpin several specialties and are often skipped.
Clinical pharmacology detail. Not just adverse effects but mechanisms, pharmacokinetics, interactions and therapeutic monitoring.
Medical ophthalmology. Only four items, but self-limited and high-yield; visual field defects and the red eye are predictable.
Palliative medicine. Opioid conversion, symptom control and syringe-driver prescribing — four dependable items many candidates never touch.
Dynamic endocrine testing. Interpreting suppression and stimulation tests, not just recognising the disease.
Haematology morphology. Blood-film and marrow interpretation, which reward pattern exposure rather than fact memorisation.
Geriatric prescribing and multimorbidity. Falls, delirium, polypharmacy and deprescribing framed for the older patient.

Format checklist: coverage, data, time

Doing questions is not the same as practising the exam. Verify that your preparation deliberately trains the format, not just the content.

You have attempted questions across every blueprint domain, in roughly blueprint proportion, within the last revision cycle.
You have completed at least a few full 100-item, three-hour timed papers rather than only short untimed sets.
Your working pace sustains roughly 1.8 minutes per item without a collapse in accuracy in the final third.
You are comfortable with best-of-five discrimination — choosing the single best answer when two options are defensible.
Because there is no negative marking, you never leave an item blank and you have a disciplined "flag, best-guess, move on" routine.

Interpretation checklist: can you read the data?

MRCP Part 1 is a written paper, but a substantial share of items hinge on interpreting a picture or a data set rather than recalling a fact. Confirm you can do each of the following at speed, from the stem alone.

ECGs: rate, rhythm, axis, blocks, ischaemia, electrolyte and channelopathy patterns.
Chest and other radiographs: the common, examinable patterns rather than subtle radiology.
Blood films and indices: haemolysis, haematinics, marrow-failure and malignancy pictures.
Laboratory trends: interpreting a panel and its trajectory, not a single flagged value.
Dynamic and functional tests: short Synacthen, dexamethasone suppression, water-deprivation and similar.
Calculations and statistics: anion gap, osmolality, corrected values, and test-performance figures.

If any of these makes you slow or uneasy, that is a content gap even when your topic knowledge is good.

Recency checklist: is your knowledge in date?

Some Part 1 answers move with the evidence. Guarding against out-of-date recall matters, and the discipline is simple: for any guidance-sensitive item, record where the answer came from and when.

You have flagged your guidance-sensitive topics — for example antithrombotics, diabetes and lipid targets, immunosuppression and infection management.
For each, you can name the source and its date, drawn from the UK stack: NICE, CKS, SIGN, the SmPC/eMC for medicines and NHS clinical content.
You have checked that your bank's explanations are current and UK-referenced, not inherited from an older edition.
You have noted the jurisdiction of any source, so that non-UK guidance is not silently carried into a UK exam.

Performance checklist: prove it on unseen questions

This is the row candidates skip, and it is the one that decides the outcome. A percentage earned on a bank you have partly seen before is inflated by memory of the items, not by mastery of the content — the reasoning is set out in why your Q-bank percentage is not your exam score. To measure readiness you need a clean signal.

You have sat at least one fresh, unseen mixed block under timed conditions, from a source you have not been revising on.
Your unseen score is stable across two or three such blocks, not a single lucky run.
Your speed holds up: accuracy in the last 25 items is not materially worse than the first 25.
You have hunted your high-confidence errors — the items you were sure of and got wrong — because these are the ones that lose marks you cannot see coming.
You have checked retention, re-testing topics revised weeks ago rather than only fresh ones.
You have calibrated against the official MRCP(UK) Part 1 sample questions, which set the tone and difficulty the examiners intend.

Using a second, unseen bank purely as your measurement instrument is the core of the two-Q-bank rule: revise on one bank, measure on another so your headline number is not contaminated by recall.

Worked example: Priya, four weeks out

Priya has completed 88% of her main bank with an overall first-attempt average of 74%, and she is asking whether she can stop doing new questions. On the percentage alone, she looks ready. Her coverage matrix tells a different story.

Cardiology shows 205 questions attempted at 79%: genuinely strong. Respiratory and gastroenterology look similar. But clinical sciences shows only 34 questions attempted at 58%, statistics within it barely touched; medical ophthalmology and palliative medicine each show zero attempted; and endocrinology, while at 71%, reveals on inspection that her errors cluster entirely in dynamic-function-test interpretation. Her 74% is a real average over an unrepresentative sample — she has over-practised her strengths and under-sampled a quarter of the marks.

She then sits a fresh, unseen, timed 100-item block and scores 66% — eight points below her bank average, with accuracy in the final 20 items dropping as she rushed. The gap between 74% and 66% is the seen-question inflation the percentage article describes. Priya's correct decision is not "keep doing new cardiology questions." It is to attempt targeted blocks in clinical sciences, statistics, ophthalmology and palliative medicine, drill dynamic-test interpretation, and re-measure on unseen material. New questions are still useful — but only aimed at the confirmed gaps.

The stop-or-continue decision tree

Use the measured gap, not the calendar or your mood, to choose the next activity.

Keep doing new questions where the matrix shows a high-weight domain with low questions-attempted. You have not yet sampled enough to know if you can do it.
Consolidate rather than expand where attempted numbers are healthy but first-attempt accuracy is amber. More new questions will not fix a topic you half-know; focused review and re-testing will.
Simulate — full timed papers — when domain coverage is broadly green but your unseen or end-of-paper performance is shaky. The gap is stamina and pacing, not knowledge.
Seek teaching or a written resource where a domain stays red despite repeated attempts. Doing the same questions again rarely rescues a genuinely misunderstood topic; a different explanation will.
Rest and protect retention when every row is green and your unseen scores are stable. Past that point, more volume mostly adds fatigue, and your job is to hold the line to exam day.

Your one-page MRCP Part 1 content-gap checklist

Copy this and keep it beside your revision. You are ready to ease off new questions only when every box is ticked.

Frequently asked questions

How do I know whether I have covered the full MRCP Part 1 blueprint? You know it from a coverage matrix, not a percentage. List every blueprint domain, record how many questions you have actually attempted in each and your first-attempt accuracy, and flag anything high-weight but under-sampled. You have covered the blueprint when every domain has been sampled in roughly the proportion the exam uses it, your accuracy is acceptable across all of them — not just your favourites — and none of the under-practised core areas such as clinical sciences and statistics has been quietly skipped.

Can one question bank be enough for MRCP Part 1? For learning content, a single high-quality bank can carry most of your revision. For measuring readiness, one bank is not enough, because once you have seen its items your score reflects recall of those questions rather than mastery of the material. The practical answer is to revise mainly on one bank and keep a second, unseen bank purely to measure yourself on fresh questions, as the two-Q-bank rule describes. That keeps your headline number honest.

What should I measure instead of my overall Q-bank percentage for MRCP Part 1? Measure four things the percentage hides: coverage (have you sampled every domain in blueprint proportion), calibration (does your confidence match your accuracy, especially on high-confidence errors), pace-adjusted accuracy (does your score hold in the final third of a timed paper), and unseen performance (what do you score on questions you have never met). Your overall percentage can rise while all four quietly deteriorate, which is exactly why it is an unreliable stopping signal.

When should I stop doing new MRCP Part 1 questions? Stop expanding into new questions when your coverage matrix is green across all high-weight domains, your unseen timed scores are stable across two or more blocks, and your remaining errors are careless rather than conceptual. Before that point, keep doing new questions but aim them at confirmed gaps rather than your comfort zone. After it, additional volume mostly adds fatigue; your time is better spent on consolidation, retention and rest.

Which MRCP Part 1 resource should I use for my weakest component? Match the resource to the type of weakness. If the gap is data interpretation — ECGs, films, imaging — use image-rich banks and pattern repetition. If it is clinical sciences or statistics, a concise written text plus targeted questions usually beats more mixed practice. If it is a specialty you keep getting wrong despite attempts, a focused review resource or teaching will help more than the same questions again. Whatever you choose, confirm it is current and UK-referenced, and compare options honestly on the iatroX comparison hub rather than by reputation alone.

Editorial notes and references

Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Exam format and blueprint figures reflect the published MRCP(UK) Part 1 structure at that date; verify the current blueprint and sample questions on the official MRCP(UK) site before relying on any specific number, as arrangements are periodically revised. Any third-party question counts or features referred to here are vendor-reported and should be confirmed on the relevant product page.

Disclosure: iatroX operates its own question bank and clinical-knowledge tools, so it competes with the resources discussed here. This article is written as an exam-level method rather than a product pitch, and iatroX is positioned only for the job it is suited to — providing fresh, unseen questions for baseline and readiness measurement — not as a replacement for your primary revision bank or for clinical teaching. Corrections are welcome through the feedback route on iatrox.com. This is the exam-level hub for the MRCP Part 1 content-gap intent; companion checklists exist for MSRA and UKMLA.

References: MRCP(UK) — Part 1 format, blueprint and sample questions (mrcpuk.org); NICE, CKS, SIGN, SmPC/eMC and NHS content for guideline currency; iatroX, why your Q-bank percentage is not your exam score and the completion-is-not-coverage blueprint-matrix method.

Complete a fresh MRCP Part 1 baseline in iatroX →