Around half pass. For UK trainees, often far fewer.
Across the last three MRCP Part 1 diets, roughly one candidate in two passed. UK trainees sitting for the first time did worse each time: 49.2%, then 41.9%, then 27.1%.
MRCP Part 1 pass rates for the last three diets: 49.6% of all candidates passed in September 2025, 48.7% in January 2026 and 50.7% in May 2026. UK trainees at their first attempt passed at 49.2%, 41.9% and 27.1%. UK graduates at their first attempt passed at 74.6%, 67.2% and 73.2%. Doctors not in UK training, most of them international graduates, passed at 49.9% in September 2025 and 49.5% in January 2026, and at 54.8% on first attempts across 2025.
Pass mark: a scaled score of 540 in September 2025, and 450 from January 2026 onwards. That meant 126 correct of 199 scored questions, then 124 of 198, then 124 of 200.
Weakest topic: oncology (September 2025), endocrinology (January 2026), genetics (May 2026). Strongest topic, all three diets: geriatric medicine.
Source: the Federation of the Royal Colleges of Physicians of the UK, per-diet performance reports and the pass-rates page. Links at the end.
I sat this exam while working in the NHS, and I've taught it for three years. The numbers below are not here to frighten you. They're here because the doctors who pass tend to be the ones who understood the exam before they revised for it. Let's do that now.
Six claims doctors make about Part 1. Which hold up?
Before the data: answer all six. Your score appears at the end, and the rest of the page explains every answer.
What share of UK trainees passed first time in May 2026?
Drag the slider to your best guess, then reveal. Most doctors guess too high.
UK trainees, first attempt, 2026/02 diet. Federation pass-rates page. Trainee status is self-declared, so treat it as indicative.
Spin through the three diets.
Each diet has its own personality. The pass mark changed in January 2026, the number of questions scored changed twice, and the weakest subject was different every time. Turn the stage to compare them.
September 2025
January 2026
May 2026
Candidate numbers, pass marks, correct-answer thresholds, medians and topic statements from each diet's Federation performance report. "Correct answers to pass" is the raw count the Federation states after equating that paper.
| Diet | Candidates | Pass mark | Correct to pass | Median score | All passed | UK trainees 1st attempt | UK graduates 1st attempt |
|---|---|---|---|---|---|---|---|
| Sept 2025 | 4,695 | 540 | 126 of 199 | 537 | 49.6% | 49.2% | 74.6% |
| Jan 2026 | 3,982 | 450 | 124 of 198 | 448 | 48.7% | 41.9% | 67.2% |
| May 2026 | 3,749 | 450 | 124 of 200 | 450 | 50.7% | 27.1% | 73.2% |
The same exam, four very different pass rates.
The headline figure hides the story. UK graduates sitting for the first time pass at roughly two in three or better. UK trainees, most of them on a full rota, pass at a rate that fell with every diet. Pick a diet and watch the bars move.
Who passed MRCP Part 1?
Federation pass-rates page, 2025/03, 2026/01 and 2026/02. "UK trainees" means doctors who declared they were in a UK training programme. "UK graduates" means doctors who qualified at a UK medical school, whatever they are doing now. "Doctors not in UK training" is the Federation's non-UK resident doctor group, mostly international graduates sitting in India, the Gulf, Pakistan and elsewhere, plus IMGs in non-training NHS posts. For May 2026 the Federation published that figure only as a chart, so it is not shown as a number. One caution on the first-attempt trainee line: the Federation does not publish how many doctors sit their first attempt in each diet, and across the whole of 2025 it was 145 UK resident doctors. A single-diet first-attempt figure therefore moves a long way on small numbers, so read the fall from 49.2% to 27.1% as a direction, not a precise measurement.
Read those two UK lines together. A UK-trained doctor who sits soon after graduation, with the basic sciences still fresh, has every chance. The same doctor three years into IMT, revising between nights, has a far thinner one. The exam hasn't changed between those two rows. The time to prepare has.
Where the pass mark actually sits.
The pass mark isn't a percentage. It's a scaled score, 540 until September 2025 and 450 from January 2026, and the Federation works backwards after each exam to find how many raw marks reach it. In all three diets, the median candidate landed within three points of the line.
Fill the grid to the pass mark
Two hundred dots, one per question. Drag until you reach what a pass needed, and see how little margin there was.
Median versus pass mark
Each curve is the spread of scores. The gold line is the pass mark. The pale line is the median candidate. Click a diet to redraw.
So "aim for the pass mark" is bad advice. When the middle of the hall sits on the line, a few questions in a subject you skipped decide whether you sit again in four months. Aim well clear of it.
Where the marks were lost.
After every diet the Federation names the topic candidates scored lowest and highest in. Three diets, three different weakest subjects. One constant at the top.
Which subject was lowest in January 2026?
Four options. Choose first, then flip the cards for all three diets.
Topic statements from each diet's Federation performance report. Across the whole of 2025, the lowest-scoring area was clinical sciences and the highest was oncology, which shows how much a single diet can swing.
Geriatric medicine topped all three diets because falls, delirium and polypharmacy are what you did on Tuesday. Genetics and immunology came bottom in May because nobody revises trinucleotide repeats between night shifts. The twist is January: endocrinology, a 14-question ward specialty, came bottom for everyone, including the doctors who passed. No subject is safe because it feels familiar.
If you are sitting from India, the Gulf or Africa, the numbers are on your side. Slightly.
Most candidates do not sit in the UK. In each of the last three diets, more than half sat at international centres, and over 3,000 of the roughly 4,000 doctors each time were not in a UK training programme. Their pass rate has been steadier than the UK trainee figure.
| Group | Sept 2025 | Jan 2026 | 2025 as a whole |
|---|---|---|---|
| Candidates at international centres | 50.3% | 48.5% | — |
| Doctors not in UK training, all attempts | 49.9% | 49.5% | 51.3% |
| Doctors not in UK training, 1st attempt | — | — | 54.8% |
| UK resident doctors in training, 1st attempt | — | — | 42.8% |
Two readings. First, the exam is the same paper everywhere, so an international graduate sitting at a centre in Delhi or Dubai has the same odds as anyone else with the same preparation. Second, roughly half of that group still fails. Candidates outside the UK tend to sit sooner after graduation, with basic sciences fresher, which helps. What hurts them is the UK-specific layer: NICE pathways, BNF prescribing, and the exact wording of UK clinical practice. Knowing the medicine and choosing the single best answer MRCP wants are not quite the same skill.
Wherever you sit, the margin is the same: a median candidate on the line, and a handful of questions deciding it.
Four things to take from three diets.
1. The pass mark reset changed the number, not the standard
450 replaced 540 after the Federation's October 2025 standard-setting exercise. Doctors read it as the exam getting easier. It didn't. The raw marks needed stayed at 124 to 126, and the overall pass rate stayed around half. Any resource still quoting 540 as the live pass mark hasn't been updated, and that's a useful test of everything else it says.
2. Your question-bank percentage is not your exam score
Banks aren't equated. Scoring 65% on a bank tells you about that bank. In May 2026 a pass needed 124 of 200, 62%, and the median candidate got exactly that. Treat a bank average as a progress marker, not a prediction.
3. First attempts are the cheapest attempts
UK graduates sitting first time pass at 67% to 75%. The same group across all attempts drops to 64% to 70%. Every resit costs a fee, four months and a chunk of confidence. The best time to prepare properly is before the first sitting, not after a fail.
4. Basic sciences, then weak areas, then everything
Clinical sciences was the weakest area across 2025 and produced the two lowest topics in May 2026. It's also 25 questions, the largest block on the paper. Start there. Then your weak specialties. Then every subject, however small, because the median candidate has no spare marks.
Half the hall had a question bank. Half the hall failed.
Almost every candidate in those three diets had a bank subscription. The pass rate still sat around 50%, and for UK trainees it fell. A question bank is a good tool. It is a poor strategy.
Here is what a bank on its own does not do, and why the data looks the way it does.
It tests recall, not understanding
You can answer fifty amyloidosis questions across three banks and still meet a version on the day you have never seen, because the exam writes a new vignette around the same concept. If you learned the answers, you are stuck. If you learned amyloidosis, you are fine. Genetics and immunology came bottom in May 2026 for exactly this reason: they are concept subjects, and banks turn them into trivia.
It has no order
A bank lets you do cardiology because you like cardiology. Nobody is telling you to start with the 25-question clinical sciences block, or to stop and revise what leaked out last month. The doctors who fail are rarely short of questions. They are short of a sequence.
It has no revision plan
Part 1 is won on the second and third pass, not the first. A bank gives you a percentage that goes up as you memorise its own questions, which feels like progress. It is not equated, it is not the exam, and in May 2026 a doctor with a 65% bank average could still land on 124 out of 200.
It does not teach exam technique
Reading the stem for "most likely", "first-line" and "next best", spotting the one decisive clue, discarding the plausible distractor. These are skills, and they are taught, not absorbed from 5,000 explanations.
It gives you no one to ask
At 11pm after a long day, the bank explanation you do not understand stays not understood. That is where the "I'll come back to it" pile starts, and that pile is where the lost marks live.
It has no finish line
Five thousand questions is a number, not a plan. Doctors on full rotas burn out trying to "complete the bank" and arrive on exam day exhausted and under-revised. One hundred and fifty well-chosen questions per system, understood and revised, beats five thousand half-finished.
None of this means drop your bank. It means put it in its place: a practice tool inside a structure, after the concept is learned and before the revision pass. That structure is what we built.
Why our doctors pass at the rates they do.
Across the 2025 and 2026 diets, between 88% and 92% of the doctors who completed our MRCP Part 1 course passed, from cohorts of 50 to 70 doctors per diet. That's our internal figure, from our own students' results, and I'd rather you held it next to the Federation's 50% than take either on trust. The gap doesn't come from a secret question bank. It comes from doing the unglamorous things the data above says matter.
We also tell doctors when not to sit. Sometimes the honest advice after a strategy call is to aim for the next diet. A deferred sitting is cheaper than a failed one.
Quick answers.
What is the MRCP Part 1 pass rate?
Across the last three diets, 49.6% of all candidates passed in September 2025, 48.7% in January 2026 and 50.7% in May 2026. UK trainees sitting for the first time passed at 49.2%, 41.9% and 27.1%, according to the Federation of the Royal Colleges of Physicians of the UK.
What is the MRCP Part 1 pass mark?
From the January 2026 diet the pass mark is a scaled score of 450, replacing 540. Because papers are equated for difficulty, the number of correct answers needed changes each diet: 126 of 199 in September 2025, 124 of 198 in January 2026 and 124 of 200 in May 2026.
How many correct answers do you need to pass MRCP Part 1?
In May 2026 a pass needed 124 correct answers out of 200 scored questions, which is 62%. In January 2026 it was 124 of 198, and in September 2025 it was 126 of 199. The figure is set after each diet, so it varies.
Which topics do candidates score lowest in?
It changes each diet: oncology in September 2025, endocrinology, diabetes and metabolic medicine in January 2026, and genetics in May 2026 (immunology among those who passed). Across 2025 as a whole, clinical sciences was the weakest area.
Which topic do candidates score highest in?
Geriatric medicine was the highest-scoring topic for all candidates in all three of the last diets.
What is the MRCP Part 1 pass rate for international doctors?
Doctors not in UK training, most of them international medical graduates sitting at centres in India, the Gulf, Pakistan and elsewhere, passed at 49.9% in September 2025 and 49.5% in January 2026 (all attempts). Across 2025 as a whole, they passed at 54.8% on first attempts, against 42.8% for UK resident doctors in training.
Why is the UK trainee pass rate lower than other groups?
The Federation's reports consistently show UK resident doctors in training passing at a lower rate than other candidates, and far lower than UK graduates sitting first time. From teaching IMT doctors, the gap looks like time and structure, not ability. A full NHS rota leaves little energy for consistent revision, and most trainees prepare without a plan built around nights and on-calls.
Dr Mustafa Baig
Internal Medicine doctor, NHS · MRCP Part 1 tutor, Intellect Medicos · MBA, University of South Wales
Dr Baig joined Intellect Medicos in 2023 as a struggling MRCP Part 1 student, passed, and joined the faculty the same year. He now designs the Part 1 curriculum and teaches on the course alongside Dr Chirag Madaan, and co-delivers the MRCP Fast Track revision workshops with Pastest in the UK.
Published by Intellect Medicos, MRCP Part 1 and Part 2 coaching for working doctors. Learning made easy.
- MRCP(UK) Part 1 September 2025 performance report, Federation of the Royal Colleges of Physicians of the UK
- MRCP(UK) Part 1 January 2026 performance report
- MRCP(UK) Part 1 May 2026 performance report
- Pass rates explained, Federation (2025/03, 2026/01 and 2026/02 tables)
- Pass marks explained, Federation
- MRCP(UK) Part 1 2025 annual performance report
Intellect Medicos pass rates are internal figures from our own students' declared results and are not independently audited. Federation figures are reproduced as published; confirm current details for your own diet on the Federation website.