This is for candidates watching their BMJ OnExamination percentage climb ahead of the MRCPCH Foundation of Practice (FOP) paper and wondering whether it means they are ready. The principal limitation to internalise now: a bank dashboard reports practice data — how you did on questions you chose, many of them seen before — and practice data is not readiness. This audit defines each number, shows the four ways it misleads, and converts the dashboard into a weekly plan instead of a false sense of security.
What BMJ OnExamination offers for MRCPCH FOP right now
Checked on 19 July 2026, all figures are vendor-reported; confirm current numbers on onexamination.com.
| Feature | What OnExamination states (vendor-reported, 19 July 2026) |
|---|---|
| Coverage | MRCPCH: FOP, syllabus-matched topics written by paediatric authors |
| Question volume | 930+ questions |
| Question selection | "Select Questions" by difficulty; revisit challenging items |
| Mocks | "Mock Tests" curated by the BMJ editorial team to recent exam themes |
| Social | "Group Learning" — daily leaderboard rounds of 10 questions |
| Access and app | Offline mobile app; free trial of 10 questions per day |
| Price | Roughly £34.99 (1 month) to £149.99 (12 months) |
One honesty note before the analysis: I did not find a claim of a proprietary adaptive-difficulty algorithm on the product page, and could not confirm which advanced metrics (percentile, predicted score, time-per-item) the current FOP dashboard exposes. This audit assumes only what the page states and tells you to verify the rest.
The FOP exam anchor
FOP is one of the first written stages of MRCPCH, frequently sat on the same day as Theory and Science (FOP morning, TAS afternoon), which makes stamina across a full day a real consideration. RCPCH sets FOP as 100 single-best-answer questions in two hours, one paper, computer-based, one mark per correct answer and no negative marking. It assesses the knowledge, understanding and clinical decision-making of a doctor entering core specialist training, and it is shared with the Diploma of Child Health. RCPCH publishes the syllabus content areas (roughly 23 domains) and official example questions, but no fixed per-domain question count — so coverage is judged by breadth, not by matching a published weighting.
Practice data is not readiness
The core error is treating a running practice average as a forecast. The dashboard summarises what you have already done, under conditions you set; the exam is a fixed, unseen, timed, mixed sample. Those are different measurements, and only the second predicts the first. Before trusting any number, ask what it actually is.
| Metric | What it is | Practice data or readiness? |
|---|---|---|
| First-attempt accuracy | Correct on first sighting | The best readiness proxy the bank offers — track this one |
| Repeat accuracy | Correct on re-attempts | Practice data; inflated by recognition |
| Percentile | Rank against other users | Practice data; a moving reference that depends on the cohort |
| Predicted score | Modelled pass-likelihood, if shown | Neither reliable nor guaranteed; verify it is even present, and never treat it as a promise |
| Coverage | Domains and items attempted | The most useful signal after first-attempt accuracy — empty domains are risk |
| Difficulty | The band a question is tagged to | Practice data; only informative if you deliberately sample hard items |
| Time per item | Seconds per question | Readiness-relevant if compared against about 72 seconds per FOP item |
Four ways practice data misleads
- Repeats: re-answering seen items measures recognition, so repeat-heavy averages drift upward without any real gain.
- Selection: whatever you filter to biases the number — weak-topic practice depresses it, comfortable practice inflates it.
- Difficulty skew: a diet of easy and medium items flatters accuracy that a hard, mixed paper will not.
- Small samples: ten-question leaderboard rounds and short sets swing wildly and prove little; readiness needs a full paper's worth.
Each of these is why a 90% home screen can sit beside a first-time-through performance in the low sixties. The fix is not to distrust the platform but to read the right number under the right conditions.
A worked illustration makes the repeat effect concrete. Suppose you have answered 1,000 items, but only 600 are unique first attempts, on which you scored 60%; the other 400 are second attempts you now get right 90% of the time. Your blended dashboard reads roughly 72% — a number that describes your memory of seen items far more than your first-time performance. The exam only ever shows you first attempts, so the 60% is the honest planning figure and the 72% is the comfortable illusion. Whenever the platform lets you, filter the analytics to first attempts only before you draw any conclusion.
The readiness test
A credible FOP readiness signal needs five conditions at once: unseen (not previously worked), timed (about 72 seconds per item), mixed (all domains, not one topic), unassisted (no notes, hints or tutor) and a sufficient sample (a full 100-question mock, not a short set). OnExamination's mock tests can supply the timed, mixed condition; the unseen condition is the one that erodes as you work the bank, which is exactly where a second, unseen source earns its place. This is the argument of the two-Q-bank rule: a primary bank for volume, a second unseen bank for measurement.
Sample size is the condition candidates most often skip. A ten-question leaderboard round tells you almost nothing — a couple of lucky guesses swing it by twenty points. A single 100-question mock gives a usable estimate of where you stand; two or three unseen mocks across the final fortnight give a trend, which is more informative than any one score. Treat one good mock as a data point and three as a direction of travel, and never re-sit the same mock and count the second attempt as evidence.
Override the comfortable defaults
Because selection and difficulty are under your control, override them deliberately:
- Force the low-volume domains the bank offers few of — genetics, metabolic, ophthalmology, community child health — until each is attempted.
- Force ethics and safeguarding items, which candidates routinely skip.
- Force a weekly hard-difficulty block, so accuracy is not propped up by easy items.
- Force at least one full timed mock per fortnight to test stamina, particularly if you are sitting FOP and TAS on the same day.
Worked dashboard example: turn analytics into quotas
Suppose the dashboard shows: blended accuracy 79%; first-attempt accuracy 61%; repeat accuracy 90%; 16 of 23 domains touched, heavy on cardiology and respiratory, light on genetics, metabolic, ethics and statistics; difficulty mostly medium; average 55 seconds per item; percentile 62nd. Here is next week's plan.
| Signal in the data | Next week's quota |
|---|---|
| First-attempt far below blended average | Trust the 61%, not the 79%; plan from it |
| Seven domains barely touched | At least 15 questions in each uncovered domain |
| Repeat accuracy inflating the blend | 25 unseen mixed questions in iatroX, unassisted |
| Difficulty skewed medium | One 30-item hard block |
| Very fast average time | One 100-question timed mock to check accuracy holds at pace |
There is deliberately no predicted pass mark here. The 55-second average is a warning, not a strength: fast-and-right is readiness, fast-and-guessing is not, and only a timed mock on unseen items tells them apart. The reasoning for refusing to turn a percentage into a prediction is set out in "Your Q-Bank Percentage Is Not Your Exam Score".
Worked example: a seven-day plan around clinical work
Assume a trainee working full time, four weeks from FOP, using OnExamination as the primary bank.
| Day | Job (about 60 min) | Tool |
|---|---|---|
| Mon | Coverage: force two uncovered domains | OnExamination (Select Questions) |
| Tue | Difficulty: 30-item hard block, review every miss | OnExamination |
| Wed | Ethics and safeguarding items, reviewed into rules | OnExamination |
| Thu | Transfer: 25 unseen mixed questions, unassisted, timed | iatroX |
| Fri | Simulate: 100-question timed mock, 2 hours | OnExamination mock |
| Sat | Feedback: misconception notes; update the coverage matrix | Your log |
| Sun | Retest: spaced recall of the week's misses | iatroX / OnExamination |
OnExamination does one job — coverage and difficulty control on a seen bank — and iatroX provides the unseen, unassisted first-attempt read on Thursday and Sunday, with no proprietary-algorithm claim on either side.
Decision checklist
- Continue with OnExamination if first-attempt accuracy on unseen items is rising and coverage is broadening.
- Supplement with an unseen bank (iatroX) when repeat accuracy is high but unseen accuracy lags.
- Switch or add a bank if whole domains are thinly stocked or your mocks do not mirror the 100-question, two-hour paper.
- Stop drilling a domain when it is broadly covered, at first-attempt strength on unseen items and at exam pace.
Bottom line
BMJ OnExamination's FOP analytics are a good management tool and a poor crystal ball. Read the first-attempt number, audit coverage, override the comfortable defaults, and take your readiness read from unseen, timed, mixed, unassisted mocks — not from the blended home-screen percentage. Add iatroX as the unseen-measurement layer, and compare banks with the iatroX comparison hub if you are still choosing.
FAQ
Is BMJ OnExamination enough for MRCPCH Foundation of Practice on its own? It can be a solid primary bank at 930+ questions, but "enough" depends on how you read it. Without a source of unseen, unassisted questions you cannot separate genuine first-time performance from recognition of seen items, so pairing it with an unseen-measurement layer — and reading the first-attempt number rather than the blend — matters more than the headline percentage.
Which MRCPCH Foundation of Practice component does BMJ OnExamination not reproduce well? The unseen condition, once you have worked the bank, and the same-day stamina of sitting FOP and TAS together — no single-paper bank reproduces a full exam day. Its mock tests can reproduce timing and mixing for one paper; confirm they match the 100-question, two-hour structure.
How many BMJ OnExamination questions should I complete per day for MRCPCH Foundation of Practice? A reviewed 40 to 60 per day around clinical work is sustainable, but the count matters less than forcing uncovered domains, sampling hard items and reviewing every miss into a corrected rule. A large unreviewed total lifts the dashboard without lifting first-attempt accuracy.
When should I stop using BMJ OnExamination and move to mixed mocks? Move to full, timed, mixed mocks once coverage is broadly complete and first-attempt accuracy on unseen items is stable — usually the final two to three weeks. Ring-fence at least one 100-question timed mock and one unseen mixed block for that window.
How should I combine BMJ OnExamination with iatroX without duplicating practice? Give each a single job: OnExamination for coverage and difficulty control, iatroX for unseen, unassisted, timed measurement. Because iatroX supplies questions you have not worked in OnExamination, the two measure different things rather than repeating items, which is what prevents duplication and keeps your readiness read honest.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; vendor figures (question counts, prices, access periods and features) are vendor-reported and change without notice — verify each on the OnExamination product page before relying on it. Disclosure: iatroX operates a UK clinical question bank that competes with BMJ OnExamination; it does not publish a dedicated MRCPCH bank, and its role here is confined to unseen-question measurement and transfer practice — a job the audited product does not claim. No proprietary-algorithm claims are made for either platform. Corrections are welcome via the feedback route on iatrox.com.
References:
- RCPCH, Theory examinations — structure and syllabi: https://www.rcpch.ac.uk/education-careers/examinations/theory/structure-syllabi
- BMJ OnExamination MRCPCH: FOP (product page): https://www.onexamination.com/products/mrcpch-fop
- iatroX, "Your Q-Bank Percentage Is Not Your Exam Score": https://www.iatrox.com/blog/qbank-percentage-not-your-exam-score
- iatroX, question-bank completion is not coverage: https://www.iatrox.com/blog/question-bank-completion-is-not-coverage-how-to-build-a-blueprint-coverage-matrix-for-any-medical-exam
- iatroX comparison hub: https://www.iatrox.com/compare
Run a fresh, timed MRCPCH Foundation of Practice block in iatroX →