Every MRCPsych candidate faces the same question: which question bank should I use? The answer has been straightforward for years — SPMM dominated the market. But in 2026, the landscape has shifted. Adaptive platforms, teaching cascades, and AI-driven blind-spot analysis have changed what a question bank can do.
As an NHS consultant psychiatrist who has taught trainees through three iterations of the MRCPsych curriculum, I have watched hundreds of candidates make the same mistake: they pick a question bank based on volume rather than intelligence. This article breaks down the options with the honest lens of someone who has seen what works in exam conditions.
What Makes a Great MRCPsych Question Bank?
Before comparing platforms, it is worth defining what a question bank should actually do. The MRCPsych is not a test of recall. Paper A tests scientific foundations (neurosciences, psychopharmacology, classification). Paper B tests critical review and clinical application. Both papers require you to apply knowledge under time pressure.
A good question bank therefore needs four things:
- Exam-fidelity questions: The style, difficulty, and weighting must mirror the real paper. Questions that are too easy or too obscure are worse than useless — they give false confidence.
- Detailed explanations: Not just why the correct answer is right, but why each distractor is wrong. The teaching is in the distractors.
- Adaptive targeting: A static bank cannot identify that you consistently confuse atypical antipsychotic side-effect profiles. An adaptive engine can.
- Performance analytics: You need domain-level breakdowns, not just a percentage score. Which section is dragging you down? Neurosciences? Psychopharmacology? Psychotherapy modalities?
SPMM — The Incumbent (6,000+ Questions)
SPMM remains the most widely used question bank among UK trainees, and for good reason. It has volume, it has brand recognition, and it is affordable at roughly £80–150 per sitting. The questions are passable and the interface is functional.
However, SPMM has not evolved significantly in the last five years. The explanations are often minimal: a single paragraph telling you why the answer is correct, but rarely explaining each distractor in detail. There is no adaptive engine, no blind-spot mapping, and no personalised remediation plan.
Candidates who use SPMM exclusively tend to plateau. They score 65–70% across the board and cannot identify which domain is holding them back. The bank gives them volume but not insight.
Best for: Candidates on a tight budget who want question volume and are prepared to supplement with other resources.
MRCPsychMentor — The Notes Platform
MRCPsychMentor is better as a revision guide than as a question bank. The notes are structured, clinically sound, and useful for last-minute consolidation. But the question bank component is limited — fewer than 500 questions, with minimal analytics.
Trainees typically use MRCPsychMentor alongside SPMM: one for notes, the other for questions. This works, but it is inefficient. You end up cross-referencing two systems rather than learning in one place.
Best for: Candidates who want a structured note set and are using a separate question bank for practice.
PassMedicine Psychiatry
PassMedicine has a strong foothold in UK medical school finals, but the psychiatry section is not deep enough for MRCPsych. The questions are oriented toward foundation-year knowledge, not the specialist depth required for Paper A neurosciences or Paper B critical review.
Best for: FY/CT1 level revision — not recommended as a primary MRCPsych resource.
PsychStar — The Adaptive Alternative
PsychStar is the newer entrant in this space and represents a fundamentally different philosophy. Rather than a static bank of questions, PsychStar uses a Bayesian adaptive engine that estimates your proficiency across every domain and selects questions specifically to probe your weakest areas.
Key differences from traditional banks:
- Teaching cascades: When you answer incorrectly, the platform does not just show you the correct answer. It triggers a multi-layer teaching cascade: concept explanation → source guideline → analogous question. You learn the underlying principle, not just the answer to one question.
- Blind-spot mapping: Your performance is broken down by domain (neurosciences, psychopharmacology, classification, psychotherapy, etc.) and by cognitive level (recall vs application vs analysis). You can see — at a glance — exactly which areas are costing you marks.
- NICE-grounded answers: Every explanation cites its source guideline or trial, so you can verify the evidence base yourself. This is particularly important for legal and ethical questions where the margin for error is zero.
- Exam simulation mode: Timed, weighted, and structured to mirror the actual paper layout. You learn to manage time as well as knowledge.
With over 4,600 questions across Paper A and Paper B, the volume is competitive with SPMM, but the adaptive layer means you spend less time on questions you already know and more time closing the gaps that would cost you marks.
Best for: Candidates who want personalised, intelligent preparation and are willing to invest in a premium adaptive platform.
Side-by-Side Comparison
| Feature | SPMM | MRCPsychMentor | PsychStar |
|---|---|---|---|
| Question count | ~6,000 | ~500 | 4,600+ |
| Adaptive engine | No | No | Yes (Bayesian) |
| Distractor explanations | Minimal | Limited | Full per-option |
| Blind-spot mapping | No | No | Yes |
| Teaching cascades | No | No | Yes |
| NICE/NHS-grounded | Partial | Partial | Yes (cited) |
| Paper A + B coverage | Yes | Yes | Yes |
| Price (monthly) | ~£80–150 flat | ~£20–50 flat | £29 per paper |
How to Choose
Your choice of question bank should match your stage of preparation:
- Early preparation (6+ months out): A traditional bank for volume is reasonable, but supplement it with a tool that identifies your blind spots. The risk of waiting too long to assess your weaknesses is that you spend six months reinforcing what you already know.
- Mid-preparation (3–6 months): This is where adaptive learning adds the most value. You have enough baseline knowledge for the engine to map your profile accurately. Targeted remediation at this stage can lift your score by 10–15%.
- Final revision (0–3 months): Exam simulation and rapid fire on weak domains. A static bank cannot provide this level of specificity.
Final Verdict
SPMM remains the default choice for volume and price. But if you are serious about passing on your first attempt — and most trainees are, given the competition for training numbers — you need more than volume. You need intelligence. The platforms that combine exam-fidelity questions with adaptive targeting are the ones that will define the next generation of MRCPsych preparation.
PsychStar offers a free 5-question trial at psychstar.io/try so you can judge the quality for yourself. No signup required.