If you want a single answer to "which MSRA resource is best", the honest reply is that the question is wrong. The right resource depends on how many weeks you have, what you can spend, and which of the two papers is actually failing you. This article is a decision tree for the Multi-Specialty Recruitment Assessment: measure your profile first, then follow the branch that fits. It is written for UK specialty-training applicants and it treats iatroX as the measurement input, not the answer.
Start here: measure before you choose
Every good MSRA decision starts with a number you can trust. Before you buy anything, sit one fresh, timed, mixed Clinical Problem Solving (CPS) block of 40 to 50 unseen questions under exam pace, and one short set of Professional Dilemmas (PD) scenarios from official material. That gives you four measurable inputs the rest of this tree runs on: your blueprint-coverage gap (which domains you have barely touched), your unseen-score trend (are fresh, timed blocks rising, flat or falling over two to three weeks), your weeks-to-exam, and your dominant error type (knowledge gap, pacing, PD/SJT calibration, or careless slips). Do not skip this. Buying a resource before you have measured yourself is how candidates end up owning three overlapping banks and still failing the paper that was never their strength.
The minimum MSRA stack
Most candidates need far less than they buy. The minimum viable stack is four things, and often you can omit the last two:
- One primary CPS question bank, worked to a target rather than to 100 per cent completion.
- The official calibration material from NHS England: the applicant guidance, sample questions and the Pearson VUE demo. This is free and non-negotiable, because it is the only source that shows you the real format and the real PD scoring style.
- One teaching or reference source, and only where a domain is genuinely unfamiliar rather than merely rusty.
- One modality tool for a specific weakness, such as a Professional Dilemmas playbook if your PD calibration is off.
If you already know your medicine and just need reps and pace, the last two are noise. Adding resources you will not finish is a cost, not a hedge.
Segment yourself first
Six profiles cover almost everyone. Find yours before you look at platforms.
- First-attempt, sound foundations. You need volume, timing and blueprint breadth, not teaching. Branch: one CPS bank plus official material plus unseen measurement.
- Retaker. Your priority is diagnosing why you missed the mark last time. If your unseen-score trend was flat, the problem is usually coverage or pace, not effort. Branch: targeted domains plus timed mixed mocks, not a fresh cover-to-cover pass of a new bank.
- Specialty-training applicant under competition-ratio pressure. You are optimising rank, not just a pass. Branch: maximise CPS accuracy on high-frequency domains and stop leaking marks on PD.
- Weak foundations or long time out of acute practice. You genuinely need teaching first. Branch: a teaching or notes source for four to six weeks, then convert to retrieval.
- Strong knowledge, poor pacing. Your accuracy is fine untimed and collapses at sub-minute pace. Branch: timed blocks and mocks only; more content will not help.
- Strong CPS, weak Professional Dilemmas. The commonest silent rank-killer. Branch: PD-specific practice against official material and a structured PD approach, because your clinical bank cannot fix this.
Budget bands (verify every price on the day you buy)
Prices below are vendor-reported and were checked on 19 July 2026; confirm the current figure on each product page before purchase, because MSRA vendors change tiers between diets.
- Free or low-cost. NHS England applicant guidance, sample questions and the Pearson VUE demo cost nothing and are the format gold-standard. iatroX free UK-core practice gives you unseen, timed CPS items for measurement. Several vendors also publish free sample sets. For a candidate with sound foundations and eight-plus weeks, this plus one modest paid bank is often enough.
- One premium resource. Pick a single MSRA bank matched to your weakest component. As vendor-reported examples on 19 July 2026: Pass the MSRA's all-access bundle was £30 for one month, £60 for three and £90 for six; Revise MSRA sold one-off access at £54.97, £69.97 and £99.97 for one, three and six months; eMedica's MSRA online revision ran from £49 for one month to £129 for twelve.
- Comprehensive stack. One teaching source plus one CPS bank plus official material plus an unseen measurement bank. The guardrail here is discipline: buying two or three CPS banks that all test the same common presentations wastes money and destroys the comparability of your scores.
Time bands: what to omit, not what to add
- Under four weeks. Omit new teaching courses and any fresh cover-to-cover bank. Do timed mixed CPS blocks, the official sample questions, focused PD practice and misses only. The goal is calibration and pace, not new coverage.
- Four to twelve weeks. Work one CPS bank to a domain target, space your misses, run one timed mixed mock weekly, and drill PD. Add teaching only for a domain you genuinely do not know.
- More than twelve weeks. Spend the first four to six weeks on foundations if you need them, then shift decisively to retrieval and timed mocks. Introduce a second, unseen bank late, purely to measure transfer, not to double your reading.
The platform-to-job decision matrix
Each MSRA resource has a job it does best. Match the job to your gap.
| Resource | Best job for the MSRA | Watch-out |
|---|---|---|
| Pass the MSRA | Breadth of notes, flashcards and podcasts for spaced recall and teaching | Passive by default; needs active timed testing bolted on |
| Revise MSRA | MSRA-specific CPS volume plus a structured Professional Dilemmas approach | Confirm current counts and analytics depth on the page |
| eMedica | Live crammer teaching and exam-style mocks; human structure | Course time competes with retrieval hours |
| Passmedicine, Pastest, Quesmed | High-volume CPS banks with analytics | Verify current MSRA coverage, counts and price per page |
| MedRevisions | Not applicable — covers PLAB 1 and UKMLA AKT, not the MSRA | Do not buy for MSRA; see the MedRevisions audit |
| NHS England material + Pearson VUE demo | Format and PD-scoring calibration; free gold-standard | Finite volume — cannot carry your whole revision |
| iatroX | Unseen, timed, mixed CPS measurement; the transfer layer | Does not reproduce the official consensus-scored PD paper |
For the detailed evidence behind each platform, follow the narrow child audits rather than duplicating them here: the Pass the MSRA gap analysis, the Revise MSRA coverage audit and the eMedica format audit.
Three worked candidate profiles
Priya, first attempt, ten weeks out, strong CPS, weak PD, budget for one premium resource. Her baseline shows CPS accuracy around her target on timed blocks but PD ranking answers drifting from the official consensus. She does not need a second clinical bank. Weekly allocation: three timed CPS blocks (about 45 items each) plus misses, two PD sessions against official sample scenarios and a structured PD approach, one weekly mixed mock. Exit criteria: PD answers matching the official rationale on eight of ten scenarios and CPS trend flat-or-rising for two weeks. Cost: one premium bank plus free official material.
Sam, retaker, six weeks out, plateaued unseen score, pacing plus two weak domains. The temptation is to buy a new bank and start again. The measured problem is pace and two domains, so a fresh 8,000-item pass would waste the six weeks. Weekly allocation: daily 20-minute timed sprints to fix pace, targeted study of the two weak domains only, two full timed mocks per week for transfer. Exit criteria: sub-55-second average per CPS item and both weak domains at cohort mean on unseen items. Cost: keep the existing bank; add nothing but iatroX measurement.
Amina, IMG applicant, fourteen weeks out, weak foundations across CPS, needs structure. She genuinely needs teaching, so she runs a comprehensive stack. Weeks 1 to 6: a teaching or notes source plus one CPS bank at a gentle pace to build coverage. Weeks 7 to 12: convert to retrieval, spacing misses and adding weekly timed mocks. Weeks 13 to 14: unseen measurement only, plus PD. Exit criteria: blueprint-coverage gap closed to fewer than three under-tested domains and a rising unseen trend. Cost: comprehensive, but sequenced so the teaching stops when retrieval starts.
The decision table
| If your measured state is… | Follow this branch |
|---|---|
| Weak PD, sound CPS | PD-specific practice plus official scenarios; do not add a CPS bank |
| Flat unseen trend despite high bank percentage | Switch to timed, mixed, unseen blocks; your percentage is inflated |
| Blueprint gap in 3+ domains, 12+ weeks | Teaching or notes first, then retrieval |
| Pacing collapse under time | Timed sprints and mocks only |
| Under four weeks, sound foundations | Official material plus mixed mocks plus misses; omit new content |
| Considering MedRevisions for MSRA | Stop — it does not cover the MSRA |
Reading your results: three mistakes this tree is designed to stop
First, chasing completion instead of coverage. Finishing 100 per cent of a bank tells you nothing if the bank under-samples your weak domains; a blueprint-coverage matrix tells you what is actually untested. Second, trusting your bank percentage as a score. It is a familiarity-inflated in-app metric, not an exam prediction; your Q-bank percentage is not your exam score. Third, buying a second overlapping bank for novelty or reassurance. If two banks test the same common presentations, you have doubled your cost and destroyed the comparability that made a second bank useful in the first place.
Evidence hierarchy
Weight your sources in this order. Official NHS England material first, for format and PD scoring. Primary UK guidance next (NICE, CKS, SIGN and the SmPC/eMC), for clinical content. Vendor pages third, for product facts, always labelled as vendor-reported. Independent candidate experience last, for usability signals only, never for coverage claims. When two sources disagree, the official body wins.
Bottom line
There is no universal best MSRA resource; there is a resource that fits your weeks, your budget and your measured weakest component. Measure first, buy the minimum stack that closes your specific gap, keep official material at the centre, and use an unseen bank to check that your learning is transferring. The decision tree above is the framework; iatroX is where you generate the baseline that tells you which branch to take.
Frequently asked questions
How do I know whether I have covered the full MSRA blueprint? You cannot tell from a completion bar. Build a coverage matrix that lists the MSRA clinical domains down one axis and, for each, records how many unseen items you have attempted and your accuracy on them. Any domain with few attempts or below-mean accuracy is an uncovered gap, regardless of your overall percentage. The Professional Dilemmas paper needs its own row, because clinical coverage tells you nothing about SJT calibration.
Can one question bank be enough for MSRA? For many well-prepared, first-attempt candidates with sound foundations, yes — one solid CPS bank plus the official material can be sufficient. A second bank earns its place only when it does a job the first cannot: giving you unseen items to measure transfer, or covering a component your primary bank handles poorly, such as Professional Dilemmas. Adding a second bank that tests the same content is duplication, not depth.
What should I measure instead of my overall Q-bank percentage for MSRA? Measure accuracy on fresh, timed, mixed, unseen blocks, broken down by domain, and track the trend over two to three weeks. Also measure your pace (seconds per CPS item) and your PD agreement with official rationales. These predict exam-day performance far better than a cumulative in-app percentage, which is inflated by repeated exposure to items you have already seen.
When should I stop doing new MSRA questions? Stop adding new questions when your unseen, timed accuracy has been stable at or above your target for about two weeks, your pace is comfortable, and your remaining errors are careless rather than knowledge-based. At that point new volume adds little and rest and light consolidation add more. Doing questions past the point of learning is reassurance-seeking, not preparation.
Which MSRA resource should I use for my weakest component? For a weak Clinical Problem Solving paper, use a high-volume CPS bank worked against a blueprint matrix so you cover the domains you avoid. For weak Professional Dilemmas, use the official sample scenarios plus a structured PD approach — a clinical bank will not fix SJT calibration, and no third-party tool perfectly reproduces the official consensus scoring. Match the tool to the component, not to its marketing.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Prices, question counts and features quoted here are vendor-reported figures that change between exam diets; verify the current figure on each product page before you buy. Disclosure: iatroX operates a competing MSRA question bank, so this article confines iatroX's role to unseen measurement and baselining — a job the resources above do not claim — rather than ranking it against them. Corrections are welcome via the feedback route on iatrox.com.
References: NHS England, Multi-Specialty Recruitment Assessment overview and structure (medical.hee.nhs.uk); vendor product pages for Pass the MSRA, Revise MSRA, eMedica and MedRevisions (accessed 19 July 2026); iatroX, "Your Q-Bank Percentage Is Not Your Exam Score" and "Question-Bank Completion Is Not Coverage" (iatrox.com/blog).