MRCGP AKT Exam Strategy for GP Trainees

Most trainees who fail the AKT the first time were not unprepared. They did questions. They did some reading. They sat a mock or two. What went wrong is usually something more specific: they treated question volume as revision, left statistics for the final fortnight, used GP Self Test as their only benchmark, or never moved from topic-specific to random mixed practice.

The patterns of people who pass comfortably — and especially those who score 90%+ — are consistent enough to be teachable. This article synthesises those patterns, updated for the current 160-question format.

Exam Facts

The AKT is a computer-based assessment at Pearson VUE test centres, delivered four times a year: January, April, July, and October. You can sit from ST2 onwards.

From October 2025: 160 single best answer questions in 160 minutes. No negative marking. One mark per correct answer.

The pass mark is set after each sitting using the Angoff methodology. Historically it has landed between approximately 61% and 72% — the difficulty of that sitting’s paper determines where in that range it falls. You receive a scaled score: zero equals the pass mark; positive means above it. The RCGP does not publish the pass mark in advance.

With 40 fewer questions than the previous format, each item carries more statistical weight. A consistently weak domain costs more than it used to.

When to Sit

Clearing the AKT in ST2 rather than ST3 is the consistent advice from trainees who planned their training strategically. ST3 brings the SCA alongside increased portfolio requirements and a full clinical caseload. The cognitive load of preparing for two MRCGP components simultaneously is real. Sitting the AKT in ST2 — and ideally in a less demanding rotation — removes that competition.

Once you know which sitting you are targeting, book the exam. Then book your study leave. Trainees who score well are almost uniformly people who did both early. A confirmed exam date turns vague preparation intentions into a schedule. Study leave blocks, once booked, are much harder for rota coordinators to move. Some deaneries allocate five days of exam study leave — claim it.

Four to six months of preparation is the commonly cited adequate lead time, particularly for trainees with family commitments, part-time rotas, or multiple responsibilities. Three months is tight but achievable if your study time is protected and structured.

The Three Domains Are Not Equal Problems

The exam splits into 80% clinical, 10% evidence-based practice, and 10% primary care organisation and management. Most trainees spend almost all of their time on clinical content and hope the rest sorts itself out. It does not.

What high scorers consistently describe is the opposite approach: treat statistics and admin as priority targets early, and finish them before the clinical domain takes over.

The reason is mechanical. Statistics and admin are bounded. The clinical domain is vast and improving slowly; the 20% domains are compact and improvable fast. One doctor who scored 89.5% overall — including 100% in statistics and 90% in admin — described deliberately clearing both domains before touching clinical content. Another who scored 94% achieved 100% in both admin and statistics. In multiple accounts, trainees who went into these sections expecting to scrape marks came out surprised by how learnable they were once properly revised.

Leaving statistics for the final two weeks is the single most common structural error in AKT prep. The domain does not respond to cramming. It responds to worked examples done repeatedly over weeks.

Clinical Knowledge: Link to Patients, Not Just Guidelines

The clinical domain covers all 32 RCGP curriculum topics. The questions test applied reasoning, not recall of isolated facts. The RCGP is explicit on this: the exam targets higher-order problem-solving.

Trainees who score highly in clinical consistently describe the same mechanism: linking revision to patients they have seen. Reading the NICE CKS entry for asthma because you just saw a step-up dilemma in clinic fixes that information differently than reading it cold. Questions on conditions you have actually managed become solvable in a way that pure reading does not produce.

The practical upshot: when you see a presentation in clinic that you are uncertain about, that is the revision. Look it up that evening, do a question block on that topic, and the two exposures compound. This is not a magic shortcut — it requires discipline to follow through on — but it is considerably more efficient than working through a curriculum in alphabetical order.

Statistics: Worked Examples Beat Memorised Formulas

The statistics domain covers: sensitivity, specificity, positive and negative predictive value, likelihood ratios, NNT, NNH, absolute and relative risk reduction, confidence intervals, p-values, study design comparison, screening test trade-offs, and interpreting graphs including spirometry, audiograms, and flow-volume curves.

Two things work: worked examples done repeatedly, and video explanations that walk through the logic step by step. Multiple high-scorers specifically credit structured statistics video content for getting to 90-100%. The value is not in the formulas — it is in seeing the same calculation approached five different ways until the reasoning becomes automatic.

One piece of practical advice that circulates among trainees: in the exam room, before you answer a single question, write down the stats facts you find hardest to hold — NNT formula, sensitivity versus specificity definitions, the 2x2 table layout. It only takes two minutes and removes the cognitive burden of trying to recall them mid-question under time pressure.

Admin and Professional Knowledge: This Domain Is Completable

The admin domain covers DVLA fitness-to-drive guidance, death certification and cremation form requirements, the Mental Capacity Act, confidentiality and disclosure duties, safeguarding thresholds, controlled drug regulations (schedules, quantities, prescription requirements), sickness certification, occupational health, and NHS organisational structures.

Trainees routinely describe encountering this content in clinical work and yet performing poorly on it in the exam. Encountering something is not the same as knowing the exact rule at the level the AKT requires. The DVLA handbook, for example, is freely available as a PDF. Most of the certification and capacity content has a defined correct process. There are no interpretation marks here: the question has a right answer, and it is findable in a sourceable document.

Treat this domain as a completable knowledge set, not an atmospheric background. Work through it methodically. The trainees who score 90%+ in admin are not more intelligent; they are people who decided it was worth their time and spent it.

Question Banks: Calibration Matters

The consensus across Reddit and trainee experience posts is fairly consistent on difficulty:

PassMedicine is harder than the real exam. Scoring 55-65% on PassMed questions does not mean you are heading for a fail. The explanations are thorough and link to guidelines, which makes it valuable for learning — but do not use your PassMed mock percentage as a direct pass/fail predictor.

GP Self Test is closer to the actual AKT format and somewhat easier. A benchmark of 70% or above on GP Self Test is a reasonable indicator you are near passing range. Aiming for 80%+ gives a comfortable buffer.

The main traps with question banks are using only one, and using them in a mode that inflates your apparent score. If you are doing cardiology questions in cardiology mode, and you know it is a cardiology session, your scores will run artificially high — there is only one plausible specialty for any answer that makes sense. After your first full pass through the curriculum, switch to random mixed questions. The discomfort of not knowing which domain you are in before reading the stem is the thing that actually prepares you for exam conditions.

The other common trap is doing questions and moving on. Review is the mechanism; answering is just the trigger. After every session, go back to the ones you got wrong or guessed and ask: did I not know the fact, did I know the topic but get the threshold wrong, or did I misread the stem? Each of those has a different fix.

Mock Exams: Weekly in the Final Month

The clearest signal from trainees who passed with high scores is mock frequency in the final four to six weeks. One trainer who scored 85% first time described running a full timed mock every week in the final month. This is not just about checking your score. It is about:

• Practising pacing (160 questions in 160 minutes — roughly one minute per question with no buffer for slow starts)
• Building tolerance for two hours and forty minutes of continuous concentration
• Identifying which topics are still generating errors at crunch time, so you can focus the final weeks of revision

Do not wait until the last two weeks to do your first full timed mock. By then you have no time to act on what it tells you.

Using Your Commuting Time

One pattern that appears in nearly every high-scorer account is using passive time productively. Commutes, waiting rooms, running, childcare routines — these are not wasted hours if you have audio content loaded. Several trainees specifically mention the Primary Care Knowledge Boost podcast and AKT-specific audio summaries as useful for maintaining familiarity with clinical content without requiring screen time.

This is not a substitute for deliberate study. It is a complement to it. Passive exposure to a topic you have recently read actively accelerates retention in a way that either alone does not.

A Practical Plan Structure

Adapt this to your timeline — the sequence matters more than the exact number of weeks.

Phase 1: Map and start the 20% (4–6 weeks)

Run a diagnostic pass: do mixed questions across all three domains to identify gaps. Build your error log. Begin statistics and admin revision immediately, not later.

Starting clinical revision at the same time is fine — the goal is to avoid arriving at weeks 9-10 with statistics untouched.

Phase 2: Target and deepen (4–6 weeks)

Move to weak-topic blocks based on your error log. Switch from topic-specific to random mixed questions — topic mode has served its purpose. Add timed question sets.

Run one mock per week. Note where errors are clustering: is it a specific topic, a question-reading pattern, or a specific domain?

Phase 3: Condition and consolidate (3–4 weeks)

Increase mock frequency. Work through your recurring error log — the notes you made from actual mistakes are worth more than rereading chapters at this point.

Practise active retrieval of your flash notes; reading them passively does not do the same work. Keep all three lanes active — the admin domain is particularly vulnerable to being parked and then going cold.

Final week: no new material. Timed practice only. Sleep.

Using ClinicQuest

ClinicQuest supports revision; it does not replace structured study or official RCGP resources.

The most useful workflow:

1. Use AKT Preparation for question practice with adaptive follow-up. When the system detects a weak topic — two wrong answers on the same concept — it generates a targeted remedial question rather than just moving on. That is the diagnostic mechanism that makes question review produce improvement.
2. Save the notes that keep recurring in errors to Lumina as short flash cards. The notes that help are small, specific, and built from actual mistakes — not chapter summaries.
3. After revealing an answer, use the AKT Teacher panel to ask “why not option C?” before moving to the next question.
4. If a weak area reflects a genuine training gap, turn it into a portfolio learning entry.

AI-generated questions and flash cards are revision aids. They are designed to support your study — not to replace curated learning resources or substitute for clinical reasoning developed through practice.

References

• RCGP. MRCGP: Applied Knowledge Test (AKT).
• RCGP. Introducing the AKT.
• iatroX. MRCGP AKT Pass Rates 2021–2026 — Every Sitting.
• GP Training Support. How I Passed the MRCGP AKT with 94% — Dr Rachel Foster.
• GP Training Support. MRCGP AKT Preparation Tips from a High Scorer — Dr Jehane Le Grange.
• GP Training Support. Tips and Resources — Dr Kaylea Clark (94%).
• GP Training Support. How I Passed My AKT Despite Thinking I Had Failed — Dr Mahfuza Islam.