The MRCGP Applied Knowledge Test is the written component of the Membership of the Royal College of General Practitioners — the exam that determines whether you can practise as an independent GP in the UK. It changed format in October 2025 (160 questions, down from 200) and runs four times per year. The pass rate has been stable at approximately 70-80% — but with a persistent and significant gap between UK graduates (85%) and IMGs (56%).
This guide covers the format, the data, the recurring examiner feedback, every major revision resource, and the adaptive approach that targets your specific weak areas.
What the MRCGP AKT Actually Tests
The AKT tests applied clinical reasoning in the context of UK primary care — not knowledge recall. You are a GP seeing a patient. The question asks what you do next. The correct answer is the NICE/BNF/SIGN answer for a UK general practice context.
Three domains:
- Clinical medicine: 80% of the exam. Applied clinical knowledge across the full GP curriculum — cardiovascular, respiratory, neurology, endocrine, MSK, mental health, women's health, paediatrics, dermatology, ENT, ophthalmology, renal, gastroenterology, and more.
- Evidence-based practice and statistics: 10%. Data interpretation (graphs, forest plots, funnel plots), NNT/NNH, sensitivity/specificity, study design, bias types, screening criteria.
- Health informatics and administration: 10%. Prescribing regulations, referral pathways, NHS structure, screening programmes, DVLA fitness-to-drive, sick notes, controlled drug regulations.
The weighting is critical: clinical medicine generates 8 out of every 10 marks. Candidates who invest disproportionate time in statistics or admin at the expense of clinical knowledge are making a strategic error.
The October 2025 Format Change — What It Means for You
From October 2025, the AKT consists of 160 questions in 160 minutes (previously 200 questions in 190 minutes). Question types remain the same: SBAs, EMQs, and free-text answers. One mark per correct answer. No negative marking. The domain weighting (80/10/10) is unchanged.
The RCGP states that this change does not affect content, curriculum coverage, or the passing standard. The Angoff method continues to be used. Candidates have slightly longer per question (60 seconds vs 57 seconds) — a modest but meaningful improvement for data interpretation questions that require calculation.
The reduced item count means data interpretation and admin questions are proportionally reduced in absolute number — but clinical medicine remains dominant. Adjust your mock exam length (160 questions, 160 minutes) but your core content strategy is unchanged.
AKT Sittings — Dates, Eligibility, Logistics
Four sittings per year (since 2024): January, April, July, October.
Eligibility: Any point during ST2 or ST3. RCGP membership required to apply. Application window: 5-6 days, opening 7-8 weeks before the sitting.
Fee: £470 per attempt. Maximum 4 attempts. Most deaneries will not allow training extension beyond this.
Venue: Pearson VUE test centres across the UK — over 150 centres, but popular locations fill fast. Book early.
Results: Typically 4-6 weeks post-sitting. Provided as scaled scores — zero equals the pass mark, positive means above pass mark.
2026 dates: January 2026 (application: December 2025), April 2026 (application: February-March 2026), July 2026 (application: May 2026), October 2026 (application: August 2026). Confirm exact dates on the RCGP website.
Historical Pass Rate Data (2021-2026)
The RCGP publishes cumulative annual pass rates in its Annual Reports and per-sitting data in AKT Feedback Reports.
2022/23 cumulative pass rate: 76.49% (RCGP Annual Report 2022-23). This means approximately 3 in 4 candidates passed across all sittings in that academic year.
The differential attainment gap: UK graduate pass rate: 85%. IMG pass rate: 56%. This 29-point gap has persisted across multiple years despite awareness campaigns. The causes are specific and addressable — UK guideline misalignment, NHS referral pathway unfamiliarity, and NICE-specific management thresholds that differ from international practice.
AKT 51 (April 2024): 1,693 candidates sat. One question suppressed (199 questions scored). This gives a sense of per-sitting candidate volume.
The pass mark varies by sitting — set using the Angoff method after each paper based on question difficulty. It has historically ranged from approximately 61-72% of available marks. No fixed pass mark is published before each sitting.
For the complete per-sitting data table, see the dedicated MRCGP AKT pass rate reference page.
What the Examiners Keep Saying You Get Wrong
The RCGP publishes AKT Feedback Reports after every sitting (AKT 47 through AKT 57 cover 2022-2025). These documents are free, publicly available, and contain the most valuable preparation intelligence available. The recurring themes are remarkably consistent.
Neurology: Flagged in four consecutive sittings. Symptom recognition for both acute and chronic neurological conditions. Medically unexplained symptoms. Candidates consistently underperform in neurology relative to other clinical domains.
Data interpretation: Graphs remain a persistent weak point across sittings. Candidates are described as "unprepared for the format" — meaning they have not practised reading forest plots, Kaplan-Meier curves, and statistical outputs under timed conditions.
Controlled drug prescribing: Flagged multiple times. CD schedules, prescription requirements, validity periods, and emergency supply provisions.
Prescribing in older people: Dose reductions, STOPP/START criteria, polypharmacy management, anticholinergic burden.
ECG interpretation: Both acute patterns (STEMI, AF, heart block) and chronic changes. Candidates who do not specifically practise ECG interpretation lose marks.
Pathology results: Not integrating all values presented in a clinical scenario. "Borderline" results missed — candidates focus on the obviously abnormal values and ignore subtle derangements.
Asthma management: Flagged multiple sittings. The exam tests current national guidance (BTS/SIGN/NICE NG80) — ensure your knowledge reflects the latest stepwise approach.
These feedback reports are the single highest-yield preparation resource beyond Q-bank practice. Download them from the RCGP website and use them to prioritise your revision.
The Best MRCGP AKT Revision Resources in 2026
iatroX — Free Adaptive AKT Bank
iatroX provides a free adaptive MRCGP AKT Q-bank mapped to the RCGP curriculum. The adaptive engine identifies your weakest clinical domains — the specific areas where you will gain the most marks — and targets them dynamically. If the examiner feedback says neurology is consistently underperformed, and your iatroX dashboard confirms neurology is your weakest domain, the engine concentrates your practice there.
NICE/CKS/BNF integration means every explanation is anchored to the primary sources the exam tests — not static authored text. Ask iatroX provides instant guideline reference when an answer is unclear. Performance dashboard shows domain-level proficiency — the data interpretation you need about your own preparation.
Free for all AKT candidates at iatrox.com/boards. MHRA-registered medical device.
Passmedicine AKT
Approximately 2,000 questions. Affordable (~£30-50). Good explanations and peer comparison. The most widely used standalone AKT bank. Limitations: no adaptive engine — you manage your own topic routing. Static explanations not linked to live guidelines.
Pastest AKT
Over 3,400 questions. Broader question variety. Detailed topic-level explanations. More expensive (~£80-150). No adaptive personalisation. Strong for candidates who want maximum question variety.
BMJ OnExamination AKT
Trusted brand. Some past-paper influence. Expensive (~£150/year). No adaptive engine, no guideline integration.
RCGP GP Self Test (Free)
Over 2,000 free questions from the RCGP. Format closest to the actual AKT. Limitations: limited analytics, some outdated questions, no adaptive routing. Essential as a format familiarisation tool.
Arora Medical Education AKT
Structured course with exam technique coaching, video courses, mock exams, and flashcards. The AKT Ultimate Package is comprehensive. Good for candidates who want a taught approach alongside question practice.
Data Interpretation — The AKT's Hardest Section
Data interpretation consistently underperforms across all sittings. Most candidates have not done formal statistics since their MPharm year — and the AKT tests it under time pressure.
What you need to know: Sensitivity, specificity, PPV, NPV, NNT, NNH, confidence intervals, p-values, odds ratio vs relative risk, forest plots (how to read them, what the diamond means, what crossing the line of no effect means), Kaplan-Meier survival curves, funnel plots, and hierarchy of evidence.
How to revise it: Active practice with worked examples, not passive reading. Complete 100+ data interpretation questions under timed conditions. The iatroX Q-Bank includes adaptive data interpretation questions targeted specifically at this domain.
Evidence-Based Practice — The Quick-Win Domain
Candidates who prepare specifically for EBM outperform those who do not by a disproportionate margin. The concepts are finite, learnable, and consistent across sittings.
Know: Hierarchy of evidence (systematic review → RCT → cohort → case-control → cross-sectional → case report → expert opinion), bias types (selection, information, confounding, lead-time, length-time), screening criteria (Wilson-Jungner), how to read a forest plot, and how to calculate NNT from absolute risk reduction.
Time investment: 2-3 weeks of focused preparation produces reliable scoring in this domain. This is the highest return-on-investment revision investment in the AKT.
12-Week AKT Revision Plan
Weeks 1-2: Baseline. iatroX diagnostic across all domains — identify clinical, EBM, and admin weaknesses. Download the most recent AKT Feedback Report.
Weeks 3-8: Clinical medicine by system. One system per week: cardiovascular, respiratory, neurology (prioritise — consistently flagged), endocrine, MSK, mental health, women's health, paediatrics, dermatology/ENT/ophthalmology. 30-50 questions daily from iatroX + supplementary Q-bank.
Weeks 9-10: EBM and data interpretation intensive. This domain is often neglected until the final week — prioritise it here instead. Complete 100+ data interpretation questions. Learn forest plot and Kaplan-Meier interpretation. Practise NNT/NNH calculations.
Week 11: Admin domain. Prescribing regulations (CD schedules, prescription validity), NHS referral pathways (2-week wait criteria), screening programmes (cervical, breast, bowel, AAA), DVLA fitness-to-drive rules, fitness for work certification.
Week 12: Mixed mock papers. 160 questions, 160 minutes — full exam conditions. Analyse results by domain. Target remaining weak areas with iatroX adaptive sessions.
IMGs and the AKT Differential
The 29-point gap between UK graduate (85%) and IMG (56%) pass rates is the most significant differential in UK postgraduate medical exams. The causes are specific.
UK prescribing conventions: Drug names (UK vs international), BNF dosing conventions, MHRA safety alerts, and the specific first-line choices that NICE recommends (which may differ from international guidelines).
NHS referral pathways: The GP-as-gatekeeper model, 2-week wait cancer referral criteria, and the specific thresholds that trigger urgent vs routine referral.
NICE-specific management thresholds: Hypertension targets (NICE NG136), diabetes management sequencing (NICE NG28), asthma stepwise approach (NICE NG80) — all may differ from international guidelines.
iatroX addresses this directly. Every explanation is grounded in NICE/CKS/BNF — the same sources the AKT tests. Ask iatroX provides instant UK-guideline-aligned management pathways for any clinical question. The adaptive engine identifies where your international training diverges from UK practice and targets those areas specifically.
Frequently Asked Questions
When should I sit — ST2 or ST3? Most candidates sit in ST2 to reduce pressure during ST3 (when RCA/SCA and portfolio compete for attention). However, sitting in ST3 gives more clinical experience. Both approaches are valid.
How many attempts do I get? Maximum 4 attempts. Most deaneries require training extension after 4 failures.
What is the pass mark? No fixed pass mark. Set by Angoff methodology after each sitting. Historically ranges 61-72%.
What happens if I fail? You can reattempt at the next sitting (subject to application). Review the AKT Feedback Report and your iatroX performance dashboard to identify specific weak domains before reattempting.
Start iatroX AKT bank free today.