There is no single best MRCP Part 1 resource, and any article that names one is guessing about you. The right stack depends on how many weeks you have, what you can spend and, above all, the specific way your knowledge is failing. This is a decision tree, not a ranking. It segments candidates by profile, defines a minimum viable stack and gives measurable rules for when to add, switch or stop — with an unseen baseline to choose your branch.
The exam you are actually preparing for
MRCP(UK) Part 1 is two three-hour papers, each with 100 best-of-five single-best-answer questions, and there is no negative marking — so you answer every item, including the ones you are guessing. Each paper is blueprinted the same way: clinical sciences carries the largest single share at 25 questions, followed by clinical pharmacology and therapeutics at 15; cardiology, infectious diseases, neurology, renal medicine, respiratory medicine, rheumatology, gastroenterology and endocrinology each carry 14; haematology 10; psychiatry 9; dermatology and geriatric medicine 8 each; oncology 5; medical ophthalmology 4; and palliative medicine 4. Across both papers that doubles, so clinical sciences alone is roughly a sixth of the whole exam — a domain candidates routinely under-revise because it feels abstract next to the specialties.
The design point that should drive your resource choice is this: Part 1 rewards broad, evenly distributed applied knowledge and precise data interpretation, not depth in three favourite specialties. A resource that lets you grind cardiology to 90% while clinical sciences sits untouched is not helping you pass.
Start with a profile, not a product
Before you compare banks, classify yourself. Most candidates fit one of six profiles, and each points to a different branch.
| Profile | Defining signal | What the stack must prioritise |
|---|---|---|
| First attempt, adequate runway | No prior sitting, 8+ weeks free | Breadth, calibration, an honest unseen baseline |
| Retake | A previous fail, known score report | Targeted repair of the domains that failed, fresh unseen items |
| Busy trainee revising around clinical work | Fragmented study, sub-hour blocks | Efficiency, mobile access, ruthless prioritisation |
| Weak foundations | Struggles with mechanism, not just facts | A teaching/reference layer before high-volume questions |
| Strong recall, poor pacing | Knows content, runs out of time | Timed, mixed, full-length blocks under exam pace |
| Strong recall, weak applied reasoning | Good on single facts, poor on data and clinical sciences | Data-interpretation and clinical-science question volume |
Note the last two. Part 1 has no clinical or practical component, so "weak practical performance" translates here into weak applied reasoning — the candidate who can define a condition but cannot read the biochemistry, genetics or statistics vignette that Part 1 loves to test. Naming your profile honestly is the single most valuable thing you do before spending money.
The minimum viable stack
Almost every successful Part 1 candidate needs four slots filled, and no more:
- One primary question bank — your main engine of active retrieval and the place you accumulate volume. This is non-negotiable and does most of the work.
- Official calibration material — the MRCP(UK) sample questions published by the Federation of the Royal Colleges of Physicians. This is the format gold standard: it shows you the true stem length, answer structure and difficulty the examiners write to. It is finite, so you use it to calibrate, not to grind.
- One teaching or reference source, only where needed — a textbook (for example a standard internal-medicine reference) or a video/tutorial layer. This slot is optional and belongs only to weak-foundation learners; strong candidates who add it usually do so to procrastinate.
- One modality tool, only where relevant — flashcards or a spaced-repetition deck for high-yield facts, or a data-interpretation drill set. Optional again, and easy to over-buy.
The failure mode is not under-buying; it is filling slot three and slot four twice each and calling a cluttered library a study plan. If you cannot say which single job each resource does, remove it.
Budget bands
Prices change and vary by promotion, so verify every figure on the day you buy. The bands, not the numbers, are what matter.
- Free / low-cost. The MRCP(UK) sample questions are free. iatroX runs free UK-core question banks, which makes it a natural unseen-measurement layer to sit alongside a paid primary bank. Budget banks in the UK market (for example Passmedicine) sit at the lower paid end and are strong on volume-for-money. A viable minimum stack at this band is one low-cost primary bank plus the free official questions plus a free measurement layer.
- One premium resource. A single higher-tier bank (for example Pastest or BMJ OnExamination) with fuller explanations and analytics, plus the free official questions. This is the right band for most first-attempt candidates with a normal budget.
- Comprehensive stack. A premium primary bank, a second measurement bank used under the two-bank rule, a teaching/reference source and a modality tool. This is justified for retakes and weak-foundation learners, and wasteful for a well-prepared first-timer.
Spending more does not raise your score; closing measured gaps does. Buy up a band only when you can name the specific gap the extra resource fills.
Time bands
The honest skill in exam planning is subtraction. State what you will not do.
- Under four weeks. Do not start a teaching course, do not open a second primary bank, and do not attempt full coverage of every subtopic. Run your primary bank on your weakest four to five domains, do the official questions once for calibration, and sit timed mixed blocks. Accept that some low-weight domains (medical ophthalmology, palliative medicine) get triage-level attention only.
- Four to twelve weeks. The standard window. One primary bank worked to completion across all domains, official questions at the midpoint and again in the final week, timed blocks in the last fortnight. Omit the second bank unless your unseen scores stall.
- More than twelve weeks. You have room for a teaching layer if your foundations are genuinely weak, and for a second measurement bank late on. The risk at this band is comfort-grinding your strong domains; schedule blueprint-proportional coverage from week one so clinical sciences and pharmacology are not left to the end.
The platform-to-job decision matrix
Each major resource is good at one job. Match the job to your profile rather than buying the most-recommended name. For the detailed evidence behind any one platform, see its dedicated audit; this hub deliberately does not reproduce those.
| Resource | Best job | Points to which profile |
|---|---|---|
| Large premium UK bank (e.g. Pastest) | Volume, analytics, exam-style breadth | First attempt, retake |
| Value UK bank (e.g. Passmedicine) | Cost-efficient volume and concise notes | Budget-constrained, busy trainee |
| Tutorial-integrated bank (e.g. Quesmed) | Teaching plus questions in one place | Weak foundations |
| BMJ OnExamination | Analytics and timed-mode realism | Strong recall, poor pacing |
| MRCP(UK) sample questions | Official format calibration | Every candidate, once |
| Reference text (internal medicine) | Mechanism and depth | Weak foundations only |
| iatroX MRCP Part 1 bank | Free, unseen measurement layer | Every candidate, as the second bank |
The decision rule: fill your primary-bank slot from the top three rows based on profile, always add the official questions, and use iatroX as the unseen-measurement second bank so your readiness signal is never contaminated by items you have already seen.
Three worked profiles
Priya, first attempt, ten weeks, normal budget. Profile: adequate runway, no glaring weakness. Stack: one premium primary bank plus the official questions plus a free iatroX measurement bank. Weekly pattern: 40–60 primary-bank questions a day in tutorial mode for weeks one to six, blueprint-proportional so clinical sciences and pharmacology are covered early; official questions at week five; timed mixed blocks from week seven; a fresh unseen iatroX block each Sunday to track transfer. Exit criterion: unseen block scores plateau above her target with even domain coverage, and timing is comfortable — at which point she stops adding new material.
Sam, retake, six weeks. Profile: previous fail with a score report showing weak clinical sciences and endocrinology. Stack: keep the existing primary bank but do not re-grind it front to back; instead drive it hard on the two failed domains, add the official questions for calibration, and add a second measurement bank so readiness is judged on unseen items. Weekly pattern: 60% of question time on the two weak domains, 40% mixed maintenance; unseen blocks twice a week. Exit criterion: the two failed domains reach parity with the rest on unseen items — the specific, measurable gap that caused the fail is closed.
Ade, busy trainee, twelve-plus weeks but fragmented. Profile: sub-hour study blocks around a full rota. Stack: one value bank on mobile plus the official questions plus iatroX for measurement; no teaching layer. Weekly pattern: short daily question sets on the phone, a spaced-repetition deck for pharmacology and clinical-science facts, one longer timed block at the weekend. Exit criterion: consistent unseen performance across a full weekend block, not a high tutorial-mode percentage accumulated in easy conditions.
Three mistakes this decision tree is designed to stop
The first is chasing the overall percentage. A rising tutorial-mode average, built on repeated and familiar items, is not evidence of readiness; it often just measures memory of a bank. The second is buying banks to feel productive. A third bank rarely adds coverage a second one did not; it usually duplicates items and destroys your calibration by mixing seen and unseen questions. The third is ignoring blueprint proportions — pouring time into a comfortable specialty while clinical sciences, pharmacology and the smaller domains stay thin. The decision tree exists to force profile-first, blueprint-proportional, measured choices instead.
The evidence hierarchy behind these choices
When sources disagree, rank them. Official material from the Federation comes first for anything about format, structure and standard. Primary clinical guidance (NICE, CKS, SIGN, the SmPC/eMC for medicines, NHS content) comes first for the actual medicine. Vendor pages are acceptable only for product facts — question counts, access periods, prices — and should be treated as vendor-reported and dated. Independent testing and candidate reports are useful for user experience but are the weakest tier for factual claims. Build your stack from the top of that hierarchy down.
FAQ
How do I know whether I have covered the full MRCP Part 1 blueprint? Map your question attempts against the published per-paper blueprint rather than trusting a completion bar. Build a simple matrix with every domain — clinical sciences, clinical pharmacology, each specialty, haematology, psychiatry, dermatology, geriatrics, oncology, medical ophthalmology and palliative medicine — and record your unseen accuracy in each. Coverage means every cell has been tested on fresh items and none is sitting empty or far below the rest; a bank that reports "100% complete" only tells you that you have seen its questions, not that you have covered the exam.
Can one question bank be enough for MRCP Part 1? For many first-attempt candidates with adequate time, one well-chosen primary bank worked thoroughly across all domains, plus the official sample questions, is genuinely enough. The caveat is measurement: once you have seen most of a bank, your percentage on it reflects recall, not readiness. That is the narrow job a second, unseen bank does — not to add more of the same content, but to give you an uncontaminated readiness signal. So one bank can be enough for learning; you still need at least a few unseen items for measurement.
What should I measure instead of my overall Q-bank percentage for MRCP Part 1? Measure per-domain accuracy on unseen, timed, mixed blocks, and watch the trend and the spread rather than the single headline number. Your overall percentage on a bank you have largely completed is one of the least informative figures available, because it is inflated by familiarity. Track how you perform on questions you have never seen, under exam pace, with domains interleaved — and check that your weakest domain is rising, because the exam does not let you skip it. See "Your Q-Bank Percentage Is Not Your Exam Score" for why the headline figure misleads.
When should I stop doing new MRCP Part 1 questions? Stop generating brand-new questions when your unseen, timed performance has plateaued at or above your target across every domain and further new items are no longer changing your error pattern. At that point the marginal question teaches you little, and your time is better spent re-testing your logged errors and sitting full-length timed papers to protect pacing and stamina. Stopping is a positive decision made on data, not something you drift into because you ran out of bank.
Which MRCP Part 1 resource should I use for my weakest component? Use your primary bank in targeted mode on that domain first, because retrieval practice on exam-style items is what moves a weak area, and only add a teaching or reference layer if the weakness is mechanistic rather than practice-volume. If you cannot interpret the clinical-science or data vignettes at all, a teaching source or textbook chapter comes before more questions; if you understand the concept but keep getting items wrong, more targeted unseen questions — measured separately from your main bank — are the fix. Match the tool to whether the gap is knowledge or transfer.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Product facts such as question counts, access periods and prices are vendor-reported and change frequently; verify them on the relevant product page on the day you buy. Disclosure: iatroX operates a competing MRCP Part 1 question bank, so this article confines iatroX's role to the job the other platforms do not claim — a free, unseen measurement layer for readiness — and does not present it as a replacement for a full teaching or primary-bank product. Corrections are welcome through the feedback route on iatrox.com.
References: Federation of the Royal Colleges of Physicians of the UK, MRCP(UK) Part 1 format and sample questions (thefederation.uk / mrcpuk.org); Pastest, Passmedicine, Quesmed and BMJ OnExamination product pages (vendor-reported); iatroX MRCP Part 1 bank (https://www.iatrox.com/mrcp-part-1); 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); "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); "The two-Q-bank rule" (https://www.iatrox.com/blog/the-two-q-bank-rule-how-to-add-a-second-bank-without-duplicating-questions-or-destroying-calibration).