The MRCPsych Paper B Q-Bank Content-Gap Checklist: What to Verify Before You Stop Doing New Questions

You have covered MRCPsych Paper B when you can produce evidence across the critical-review component and the clinical topics, not when a question bank reports it is finished. This checklist is for psychiatry trainees in the final weeks before a sitting who want a defensible answer to a single question: is anything left that self-selected practice has been quietly hiding? The minimum evidence is coverage of every clinical domain and the critical-appraisal load, first-attempt accuracy on unseen items, current guidance and mental-health law, and the ability to work statistics and read data displays rather than merely recognise a term — never a completion percentage.

What "covered" actually means for this paper

"Covered" is an evidential claim, not an activity log. Finishing a bank tells you that you have seen its items; it says nothing about the domains it under-samples, the statistics you avoided because they felt hostile, or the clinical guidance that has moved since the questions were written. Paper B is distinctive because it pairs broad clinical psychiatry with a substantial critical-review component — statistics, study design and evidence appraisal — and that critical-review load is exactly where candidates who are strong clinically tend to leave the largest hidden gap. This checklist converts a vague sense of readiness into tick-boxes you can either satisfy or fail honestly.

The current exam snapshot

Anchor your audit to the Royal College of Psychiatrists syllabus and the College's sample questions, which are the only authoritative statements of format and standard. Verify these details on rcpsych.ac.uk before you rely on them.

Feature	Paper B
Content	Critical review (a substantial critical-appraisal and statistics component) plus clinical topics across the psychiatric subspecialties
Question types	Multiple choice questions plus extended matching items, in a balance the College describes as approximately two-thirds MCQ and one-third EMI
Approximate length	Around 150 questions in three hours (confirm the current count and duration)
Marking	One mark per question; no negative marking
Delivery	Computer-based at Pearson VUE (test centre or online with remote proctoring)

The syllabus spans general adult, old age, child and adolescent, forensic, learning disability, substance misuse and liaison psychiatry, the psychotherapy evidence base, the organisation and delivery of services, and mental-health law — alongside the critical-review strand of research methods, statistics and evidence-based practice. The College publishes the approximate MCQ-to-EMI balance but not a fixed number of questions per domain, so treat the exact weight of the critical-review component as something to confirm against the syllabus rather than a provider's summary; historically it has been a substantial share of the paper. Critically, much of the critical-review strand demands doing — working a calculation, reading a plot — which a recognition-heavy bank under-trains.

Build a blueprint coverage table

Completion is not coverage. Build a coverage table that forces you to record what you have actually done in each domain rather than trusting an overall bar, using the completion-is-not-coverage method.

Blueprint domain	Official weight (verify on syllabus)	Questions attempted	First-attempt accuracy	Last reviewed	Confidence (1–5)
Critical appraisal and study design	Substantial
Statistics and evidence interpretation	Substantial
General adult psychiatry	Indicative
Old age psychiatry	Indicative
Child and adolescent psychiatry	Indicative
Forensic psychiatry and risk	Indicative
Learning disability	Indicative
Substance misuse and liaison	Indicative
Psychotherapy evidence base	Indicative
Mental-health law, services and policy	Indicative

The column that exposes hidden gaps is first-attempt accuracy, because it is the only figure a re-drilled bank cannot inflate. If the critical-review rows are thin or shaky, mark the confidence low and treat the component as un-covered until an unseen block says otherwise — this is the single most common Paper B blind spot.

Ten domain-level blind spots self-selected practice hides

These are the areas most likely to remain invisible if you practise only what you enjoy. Each should be reviewed against primary sources and, where clinical or legal judgement is involved, checked with a supervising psychiatrist before you call it done.

Randomised controlled trial methodology — randomisation, allocation concealment, blinding and intention-to-treat, and the biases each guards against.
Diagnostic-test statistics — sensitivity, specificity, positive and negative predictive value, and likelihood ratios, worked from a two-by-two table rather than recalled.
Effect measures — relative and absolute risk, odds ratios, numbers needed to treat, and confidence intervals versus p-values.
Systematic reviews and meta-analysis — reading a forest plot, understanding heterogeneity, and recognising publication bias on a funnel plot.
Study designs — cohort, case-control and cross-sectional designs and their characteristic strengths and biases.
Treatment-resistant illness — the evidence base for clozapine, lithium augmentation and other next-line strategies, with their monitoring.
Old age and organic psychiatry — dementia, delirium and late-life mood disorder, where guidance and evidence move.
Risk assessment and forensic frameworks — structured risk tools and the legal context around them.
Mental-health law and capacity — the relevant Acts and codes of practice, which differ by jurisdiction and are subject to reform.
Service organisation and policy — care pathways, quality frameworks and the evidence behind service models.

For statistics and appraisal, work problems from a critical-appraisal reference and real papers; for clinical and legal content, work from current NICE guidance, the relevant mental-health legislation and the College reading list rather than a provider's paraphrase.

Format checklist: statistics, critical appraisal, evidence interpretation and clinical policy

Four strands of Paper B behave differently from ordinary recall, and each needs deliberate practice rather than passive reading.

Statistics. Work the calculations by hand — sensitivity and specificity from a two-by-two table, numbers needed to treat, confidence-interval interpretation — until the method is automatic under time.
Critical appraisal. Appraise a real paper against a structured checklist, because appraisal is a performance skill; recognising the definition of allocation concealment is not the same as spotting its absence in a methods section.
Evidence interpretation. Practise reading the data displays the paper uses — forest plots, funnel plots, survival and ROC curves — so that a plotted result is as legible to you as a sentence.
Clinical policy. Keep guidance and mental-health law current, and note the jurisdiction, because policy is the most time-sensitive content in the paper.

Interpretation checklist

Paper B is the more interpretation-heavy of the two written papers, so check you can read every input it might supply. The relevant items are: statistical output and data displays (two-by-two tables, forest and funnel plots, survival curves, ROC curves and confidence-interval graphics); clinical monitoring data such as lithium and clozapine levels and ECG QTc in the context of psychotropic prescribing; and laboratory trends that drive a management decision. This is where "interpretation as applicable" is at its most literal — a large slice of the critical-review component is data-reading rather than fact recall. The test is whether, given a plot or a table, you can extract the correct conclusion under time.

Recency checklist

Clinical policy and law are the fastest-moving content in Paper B, so record the date and jurisdiction of the source you learned each of these from and re-verify anything older than your last guideline cycle:

current NICE guidance in high-yield areas such as depression, psychosis, dementia and self-harm;
mental-health legislation and any reform, which differs across England and Wales, Scotland and Northern Ireland;
capacity law and codes of practice, again jurisdiction-specific;
evolving evidence on next-line treatments and their monitoring.

Jurisdiction is not a footnote here: a fact that is correct under one nation's Act may be wrong under another's, so always tag which legal framework a point belongs to.

Performance checklist

Coverage is necessary but not sufficient; you also need evidence that the knowledge holds up under exam conditions. Confirm all five:

Unseen, timed, mixed blocks. Your headline number should come from questions you have never met, mixing critical review with clinical items, at roughly seventy seconds each — not from a bank you have drilled.
Speed. Data-interpretation items are slow; rehearse them under time so a forest plot does not cost you three clinical marks elsewhere.
High-confidence errors. Track items you were sure about and still got wrong, especially in statistics, where intuition is often misleading.
Retention. Re-test the critical-review component after a gap; appraisal skill decays without practice on fresh papers.
Official-material calibration. Sit the College's sample questions late, under timed conditions, and remember that your Q-bank percentage is not your exam score.

The stop-or-continue decision tree

With your coverage table and performance data in front of you, base the decision on the measured gap, never on days already invested.

Continue new questions where a domain — very often the critical-review strand — shows low first-attempt accuracy or thin volume.
Consolidate where accuracy is high but retention is slipping — space your reviews rather than adding material.
Simulate where domains are individually solid but you have never assembled a full, timed, mixed paper that interleaves data-interpretation with clinical items.
Seek teaching where the same statistical or legal error recurs despite review — a journal club or supervisor closes a gap solo drilling cannot.
Rest where performance is plateauing and fatigue is rising — recovery is a legitimate revision activity.

A one-page checklist you can copy

Copy this into your notes and tick honestly. You are covered only when every line is satisfied on unseen, timed material.

A worked example

The figures here are illustrative and invented; use them as a model for reading your own table, not as a benchmark. "Aisha" is a trainee five weeks out with an overall bank percentage of 76%. The table exposes the pattern: general adult and old age psychiatry sit at 83–86% first-attempt accuracy and were reviewed last week, but critical appraisal shows 55% with repeated forest-plot and confidence-interval errors, and mental-health law was last reviewed six weeks ago at 60% and predates a recent reform. The decision is not "keep grinding the whole bank." It is to continue new questions in critical appraisal, seek teaching by joining a journal club to appraise real papers, consolidate the strong clinical domains with spaced reviews, and simulate a full timed paper that interleaves statistics with clinical items. No pass prediction follows from these numbers — the value is that a comfortable headline percentage concealed the critical-review component that most often decides Paper B, and the table surfaced it.

Frequently asked questions

How do I know whether I have covered the full MRCPsych Paper B blueprint? You know when a coverage table — not a completion bar — shows adequate first-attempt accuracy across every domain on unseen items, with the critical-review rows scored as honestly as the clinical ones and a recent review date against each. The College syllabus (the clinical subspecialties, mental-health law and services, plus the critical-review strand of statistics, study design and evidence-based practice) is your row list; because exact per-domain counts are not published, treat the blueprint as a checklist of areas you must perform in, and calibrate against the official sample questions.

Can one question bank be enough for MRCPsych Paper B? One well-mapped bank can be your backbone, but a single bank creates a blind spot it cannot see, and for Paper B there is a second reason to widen: the critical-review component is a performance skill best built by appraising real papers, not only by answering recognition items. Use the two-Q-bank rule — a primary bank for volume and a second, unseen bank for measurement — plus the College's sample questions and hands-on appraisal of published studies.

What should I measure instead of my overall Q-bank percentage for MRCPsych Paper B? Measure first-attempt accuracy on unseen items by domain, with the critical-review strand tracked separately; your speed on data-interpretation items; retention of appraisal skill after a gap; and your high-confidence error rate in statistics. An aggregate percentage on a re-drilled bank hides the critical-review component that most often decides this paper, because strong clinical scores can mask a weak statistics score inside the same average. Component-level, unseen, timed data is the honest signal.

When should I stop doing new MRCPsych Paper B questions? Stop adding new questions in a domain when its first-attempt accuracy on unseen items is comfortably above target and stable across a retention gap. For the critical-review strand, add a second test: you should be able to appraise a fresh paper and interpret its plots without the bank's scaffolding. Keep doing new questions where accuracy is low or volume is thin, and in the final week shift towards full, timed, mixed papers and the College sample questions, reserving new questions to patch a confirmed gap.

Which MRCPsych Paper B resource should I use for my weakest component? Match the resource to the failure mode. If critical appraisal or statistics is weak, combine a bank that explains the working with hands-on appraisal of real papers in a journal club; the iatroX Paper B bank pairs blueprint-mapped items with a Socratic tutor that walks through the reasoning rather than just marking the answer, which helps for the statistics strand — but a bank does not replace appraising live studies. If clinical or legal domains are weak, work from current NICE guidance and the relevant mental-health legislation. You can compare options on the iatroX comparison hub.

Editorial notes and references

Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026. Exam format is taken from the Royal College of Psychiatrists syllabus and marking-scheme materials; the College describes both written papers as approximately two-thirds MCQ and one-third EMI, and the approximate 150-question, three-hour structure, together with the weight of the critical-review component, should be confirmed on rcpsych.ac.uk, as counts and rules can change. Any figures attributed to iatroX (for example bank size) are vendor-reported and not independently audited — verify the current count on the product page. Disclosure: iatroX operates an MRCPsych Paper B question bank and therefore competes with other psychiatry banks; this article confines iatroX's role to the written-knowledge and unseen-measurement jobs a single resource cannot do for itself, and it does not replace hands-on critical appraisal of real papers. iatroX is the knowledge and unseen-MCQ layer only — it is not a CASC simulator, and the clinical CASC is a separate examination. Corrections are welcome via the feedback route on iatrox.com. References: the Royal College of Psychiatrists syllabus, sample questions and marking scheme (rcpsych.ac.uk); current NICE guidance and the relevant mental-health legislation for clinical and legal content; and, on iatroX, the Paper B bank, the comparison hub, the completion-is-not-coverage and two-Q-bank pillars, and "Your Q-Bank Percentage Is Not Your Exam Score."

Complete a fresh MRCPsych Paper B baseline in iatroX →