This audit is for the psychiatry trainee preparing MRCPsych Paper B who wants to know what BMJ OnExamination's so-called adaptive engine is actually optimising for. The short answer, from the platform's own pages, is engagement and peer position through user-selected difficulty, curated high-impact items and mock tests — not a proprietary difficulty-adapting algorithm, and not exam-standard readiness. It addresses the written Paper B, including its critical-review third; it does not rehearse the CASC.
What BMJ OnExamination offers for MRCPsych Paper B right now
The figures below are vendor-reported, from the BMJ OnExamination MRCPsych Paper B product and platform pages as observed on 19 July 2026. Confirm the current numbers on the product page before you buy.
| Item | Vendor-reported detail (19 July 2026) |
|---|---|
| Bank size | Over 220 best-answer MCQs and over 50 extended matching items (270+ total) — smaller than the Paper A bank |
| Question format | MCQs and EMIs, matching the exam's two formats |
| Difficulty control | User-selected "Select Questions" difficulty and topic filtering |
| Analytics | Performance graphs benchmarked against peers, high-impact ranking, "revision plans to turn weaknesses into strengths" and a daily question; no published percentile or predicted pass score |
| Mock tests | Timed mock tests curated by BMJ editorial doctors |
| Explanations | Detailed answer explanations |
| Access and price | Subscriptions from 1 month (£29.99) to 12 months (£139.99), in monthly tiers — verify current pricing |
| Mobile and trial | Mobile app with offline access; daily free questions |
| Exam components | Written Paper B only — no CASC simulation |
What the "adaptive engine" is actually optimising
Because this article's premise is the engine, be precise about it. On the pages we checked, BMJ OnExamination does not describe a proprietary algorithm that measures your ability and auto-selects your next item by difficulty. What it optimises for is engagement and comparison: you set the difficulty and topic filter; a "high impact" set curated by editors and peers is pushed at you; a daily question keeps you returning; leaderboards benchmark you against other subscribers; and "revision plans" (vendor-reported wording) nudge you toward weak areas you flag. Those are useful study aids, but they optimise time-on-task and relative standing, not an exam-standard estimate — and they can be steered away from the material you find hardest. For Paper B that hardest material is usually the critical-review third, which anxious candidates avoid and a smaller bank may sample thinly. Read the engine as curation you drive, not adaptation done to you, and verify the current feature language yourself.
The exam this dashboard is trying to predict
MRCPsych Paper B is a computer-based written paper at Pearson VUE — approximately 150 questions in three hours (verify the current count and duration on rcpsych.ac.uk), mixing best-answer MCQs and EMIs in an approximately two-thirds MCQ, one-third EMI balance confirmed by the College. Paper B pairs clinical topics — the organisation and delivery of care, psychopharmacology in practice and the clinical subspecialties — with a substantial critical-review component: research methods, statistics, evidence appraisal and the interpretation of study results. That critical-review third is what distinguishes Paper B, and it is where a candidate who has drilled clinical MCQs can still come unstuck. The clinical CASC (OSCE) is separate; neither BMJ OnExamination nor iatroX simulates it.
Every metric on the dashboard, defined
First-attempt accuracy is the closest proxy for unseen ability, within attempted topics and only if you answer before reading the explanation. Repeat accuracy rises through recognition, so a near-ceiling repeat score means little; the gap between them is your memorising-versus-understanding signal. Percentile and peer benchmark is a graph against other subscribers — directional, not a national percentile and not a pass probability. There is no published predicted score. Coverage for Paper B must be read across two axes: the clinical topics and the critical-review third, the latter easily neglected in a 270-item bank. Difficulty is a filter you set, so it shapes your percentage. Time per item is roughly 72 seconds (150 questions in 180 minutes), and critical-review items — reading a results table, working an estimate — can eat that budget, so time them.
Why the headline percentage is not your exam score
Your Paper B percentage is scored on the items you filtered to, in the modes you chose. If the critical-review items are a minority of a small bank and you rarely select them, a healthy clinical-MCQ percentage overstates your readiness for a paper where statistics and appraisal carry real weight. That is the argument of "Your Q-Bank Percentage Is Not Your Exam Score": the denominator and the topic mix are doing the talking. Treat the headline as a study metric and generate the readiness signal from unseen, mixed blocks that deliberately include critical review.
The blueprint audit BMJ OnExamination will not run for you
Split your attempts into clinical topics versus critical review, and check the EMI share. The College publishes the critical-review content and a reading list rather than fixed per-topic counts, so the aim is balance, not an exact percentage — verify on rcpsych.ac.uk.
| Paper B area | Attempted (%) | First-attempt accuracy | Flag |
|---|---|---|---|
| Organisation and delivery of care | |||
| Psychopharmacology in practice | |||
| Clinical subspecialties | |||
| Critical review — study design and bias | |||
| Critical review — statistics and results interpretation | |||
| EMI-format items (across areas) |
Fill it from your own data using the matrix in "Question-bank completion is not coverage". A thin or low-scoring critical-review row is the highest-priority override for Paper B.
The readiness test — five conditions for a credible signal
Require all five: the block is unseen; timed at roughly 72 seconds per item; mixed across clinical and critical-review content rather than filtered; sat with no assistance; and drawn from a large enough sample. Add the Paper B refinement: the unseen block must contain a representative share of critical-review and EMI items, or it will flatter you on precisely the content that decides this paper.
Algorithm override rules
Force these onto a weekly quota: critical-review essentials — study designs, bias and confounding, sensitivity and specificity, likelihood ratios, number needed to treat, forest plots and meta-analysis, screening statistics; EMIs across topics; and psychopharmacology in practice, taking any drug detail from the SmPC/electronic medicines compendium rather than a memorised figure. Because the bank is comparatively small, supplement the critical-review third with the College's reading list and worked examples rather than relying on bank volume alone. A rising overall percentage must not be allowed to hide the critical-review content.
Worked dashboard example: from analytics to next week's quotas
Take a candidate five weeks out: overall first-attempt 72%; clinical subspecialties 77%; psychopharmacology 74%; organisation of care 70%; critical-review study design 58%; critical-review statistics 51%; and EMIs under-attempted. The pattern is clear — the clinical topics are carrying the number and the critical-review third is exposed.
| Area / format | Dashboard reading | Next-week quota | Mode |
|---|---|---|---|
| Critical review — statistics | 51% first-attempt | 60 items plus worked examples | Topic-filtered, timed |
| Critical review — study design | 58% first-attempt | 50 items | Topic-filtered |
| EMI-format items | Under-sampled | 40 EMIs across areas | Mixed, timed |
| Whole-blueprint check | Not measured cold | one full-length timed mixed mock | Unseen, critical review and EMIs included |
No pass prediction attaches; the plan targets the critical-review gap and forces EMIs into the week.
Seven-day pattern: BMJ OnExamination for one job, iatroX for unseen transfer
For a trainee revising around clinical work, use BMJ OnExamination for one job — building and repairing written recall — and iatroX for unseen transfer measurement. No proprietary-algorithm claims are made for either tool, and neither is a CASC simulator.
- Day 1 (Mon) — BMJ OnExamination: 30 critical-review statistics items plus a worked example on each miss.
- Day 2 (Tue) — 30 critical-review study-design items; tag error types.
- Day 3 (Wed) — 30 clinical items plus spaced re-test of Monday's misses; then a 30-item unseen, timed, mixed iatroX MRCPsych Paper B block including critical review and EMIs, first-attempt recorded by area.
- Day 4 (Thu) — a dedicated EMI set across areas, timed.
- Day 5 (Fri) — one timed BMJ OnExamination mock under full conditions.
- Day 6 (Sat) — review the mock; a short critical-appraisal read from the College reading list.
- Day 7 (Sun) — a full-length unseen, timed, mixed iatroX block; compare with Wednesday.
This is the two-bank pattern from "The two-Q-bank rule": learn on one bank, measure on a different, unseen one. If you also sit Paper A, the companion Paper A audit covers the scientific-basis domains.
Decision checklist: continue, supplement, switch or stop
| Signal (measurable, not sunk cost) | Action |
|---|---|
| First-attempt rising across clinical and critical-review content; EMIs well practised | Continue |
| Clinical percentage strong but critical review under 60% or under-attempted, or the bank feels thin on statistics | Supplement — add unseen mixed blocks with critical review plus the College reading list |
| Coverage broad and first-attempt stable, but pacing over 72 seconds on results-interpretation items | Switch emphasis to timed mixed mocks |
| Repeat accuracy at the ceiling, first-attempt plateaued, unseen blocks confirm the standard including critical review | Stop drilling; taper |
Bottom line
BMJ OnExamination's Paper B "adaptive engine" is best understood as curation you drive — filters, high-impact sets, mocks, a daily question and peer graphs — optimising engagement and relative standing rather than an exam-standard estimate, with no proprietary difficulty-adapting algorithm on show. It is a competent Paper B bank, though a comparatively small one at 270+ items, and its structural risk is the critical-review third: statistics and appraisal that anxious candidates avoid and a small pool may sample thinly. Use it to learn and repair; supplement the critical-review content with the College's material; and measure readiness on unseen, timed, mixed blocks that include critical review and EMIs. It is a written layer only — the CASC needs separate preparation.
Frequently asked questions
Is BMJ OnExamination enough for MRCPsych Paper B on its own? Possibly as a main bank for the clinical topics, but the 270+ item pool (vendor-reported, 19 July 2026) is comparatively small and the critical-review third — statistics and appraisal — may need supplementing from the College's reading list and worked examples. It also does not measure you on unseen, mixed, timed blocks once you recognise items, and it does nothing for the CASC. Treat it as one contribution to the written paper, not a complete preparation, and verify the current count on the product page.
Which MRCPsych Paper B component does BMJ OnExamination not reproduce well? The CASC, which is a separate OSCE, and — within the written paper — the critical-review third if you let filters steer you toward clinical MCQs. Statistics, study design and results interpretation carry real weight in Paper B, and a small bank can sample them thinly, so confirm your critical-review coverage and pace those items deliberately rather than trusting a clinical-topic percentage.
How many BMJ OnExamination questions should I complete per day for MRCPsych Paper B? Balance beats volume; a sustainable pattern for a working trainee is roughly 30–50 first-attempt items on weekdays with every miss analysed, deliberately weighted toward critical review and EMIs, plus a weekend mock. Given the bank's size, pair the questions with worked statistical examples rather than expecting volume alone to build appraisal skill, and protect one weekly timed, unseen, mixed block for measurement.
When should I stop using BMJ OnExamination and move to mixed mocks? When your coverage is broad across clinical and critical-review content, your first-attempt accuracy is stable including on statistics items, and recognition is creeping into repeats — not at bank completion. Because the pool is small you may reach that point relatively early, at which stage shift to full-length timed mixed mocks and the College's sample questions, keeping BMJ OnExamination for targeted critical-review drilling.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Give each a single job. Use BMJ OnExamination to learn and repair — filters, explanations, spaced review — and use iatroX only as the unseen, timed, mixed measurement layer, ensuring the measurement block carries critical-review and EMI items. Because the item pools differ, an iatroX block is genuinely unseen and does not repeat BMJ OnExamination questions, so your calibration stays honest. Neither is a CASC simulator, so plan clinical-skills practice separately.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; BMJ OnExamination question counts, prices and features are vendor-reported as of that date and change without notice — verify the current figures on the product page before purchase. Disclosure: iatroX operates a competing MRCPsych Paper B question bank, so this audit confines iatroX's role to the unseen-measurement job BMJ OnExamination does not claim to perform, and takes no position on price; neither product is a CASC simulator. Corrections are welcome via the feedback route on iatrox.com.
References: Royal College of Psychiatrists — MRCPsych exams, Paper A and B marking scheme, Paper B critical-review content and reading list (rcpsych.ac.uk; verify the current count, duration and syllabus there); BMJ OnExamination MRCPsych Paper B product page (onexamination.com; vendor-reported figures); iatroX MRCPsych Paper B bank (https://www.iatrox.com/mrcpsych-paper-b); iatroX comparison hub (https://www.iatrox.com/compare); "Your Q-Bank Percentage Is Not Your Exam Score" (https://www.iatrox.com/blog/qbank-percentage-not-your-exam-score).