The Applied Knowledge Test is 150 questions covering the breadth of Australian general practice, and it uses two question formats: single best answer items with five options, and extended matching items with a longer option list designed to force sharper discrimination between closely related choices. The pass mark is set by a modified Angoff process and varies between cycles, and there is no negative marking. What the format does not tell you, and what decides the exam for a substantial number of candidates, is that the standard being assessed is not general medical competence. It is competence as an independent Australian GP, and those are different things. Confirm current timing and format with the RACGP before you plan.
Key takeaways
- The AKT is 150 questions, using both single best answer and extended matching formats.
- The pass mark is set by modified Angoff and varies by cycle, so chasing a fixed percentage is not useful.
- There is no negative marking, so leaving a question blank is a mark surrendered for nothing.
- The exam tests Australian guidance and the Australian health system, not general medicine.
- The organisational and health-system content is finite, examinable, and routinely neglected.
What the exam is actually assessing
Read the RACGP's own framing and it is unambiguous: the AKT assesses whether you are ready for unsupervised general practice in Australia.
That word "Australia" is doing a lot of work, and it is the word candidates skim.
The clinical medicine is largely universal, and if you are a competent doctor you already hold most of it. What is not universal is the guidance that defines the correct answer, the medicines that are actually used and funded, the referral pathways that exist, the screening programmes that run, and the system within which all of it happens.
An internationally trained candidate can reason impeccably from excellent medicine to an answer that is wrong for this exam, and that is not a knowledge failure. It is a calibration failure, and revising harder will not fix it.
Where Australian practice actually diverges
The divergence concentrates in the places where the health system, rather than the disease, determines the answer.
Therapeutic guidance. Australian antibiotic guidance is specific, it is authoritative, and it is what the exam expects. Empirical choices, durations and the stewardship logic behind them are Australian, and answers drawn from another country's guidance will be plausible and incorrect.
Prescribing and subsidy. What is available, what is subsidised, what requires authority, and the practical constraints on prescribing in Australian primary care. This is health-system knowledge and it cannot be inferred from clinical competence.
Preventive activities. Which screening programmes exist, who is eligible, at what age, at what interval. This is entirely national and entirely learnable, and it is one of the highest-yield domains in the exam because it is pure recall and most candidates half-know it.
Cardiovascular risk assessment. The tool, the thresholds and the resulting management are Australian, and a candidate applying a different country's calculator will reason correctly to the wrong conclusion.
Referral and access. What a GP manages, what is referred, to whom, and under what arrangements. This is structural knowledge and it is examined.
The domain nobody revises
Beyond clinical medicine, the AKT tests the organisational, ethical and health-system dimensions of Australian general practice, and this content has a familiar profile: finite, learnable, examinable, and reliably postponed.
Professional and ethical obligations. Confidentiality and its limits. Consent. Mandatory reporting duties. Quality and safety in practice. The structure of Australian primary care and how it is funded and organised. The practical business of general practice.
None of it feels like medicine, which is exactly why candidates leave it until the final fortnight and why it decays before the exam. Give it short, frequent sessions from the beginning, and use spaced retrieval rather than reading, because these are arbitrary, non-derivable rules that will not survive a single pass.
The two question formats need different technique
The AKT uses single best answer items and extended matching items, and they reward slightly different habits.
Single best answer, with five options, is the familiar format. Read the stem, decide what you think before you look at the options, and then find it. The discipline is to form your answer before the options can anchor you.
Extended matching, with a longer option list, is deliberately constructed to punish pattern-matching. The list will contain several closely related choices, and the discrimination between them is the point of the item. When several options look reasonable, that is not a badly written question. That is the question.
The technique for these: work the case fully, decide what the discriminating feature is, and only then look at the list. Scanning the options first and choosing the one that feels familiar is exactly the behaviour the format exists to catch.
Answer everything
A brief but genuinely valuable point.
There is no negative marking. An incorrect answer costs you nothing that a blank does not. That means a blank answer is a mark given away for free, and candidates still leave them.
If time is short, ensure every question has an answer before you go back to refine the ones you flagged. Eliminate what you can and guess between the rest.
Do not chase a percentage
The pass mark is set for each cycle by a modified Angoff process, which means it moves with the difficulty of the paper while the standard stays constant.
So a colleague's account of the percentage they needed is not information you can use, and aiming to scrape a threshold that shifts is a poor strategy. Build a comfortable margin of genuine competence instead, and measure yourself on first-attempt accuracy on unseen questions rather than on a dashboard figure inflated by repeats.
Prepare with the KFP, not separately
The final planning point.
The AKT and the KFP test different capabilities and share almost all their clinical content, which means preparing for them separately duplicates the work.
Build the content once, across the breadth of Australian general practice, then practise it in two modes: as knowledge questions for the AKT and as decision scenarios for the KFP. The KFP's format changed substantially at the 2025.2 sitting, and we cover what that means in the new KFP format.
Where iatroX fits
iatroX's RACGP bank targets the domains where you are genuinely weak rather than the ones you enjoy, with spaced repetition to hold the preventive schedules, thresholds and organisational rules that decay fastest, and missed questions can be opened in the Socratic Tutor, which asks you to reason before it explains. That reasoning step is precisely what surfaces an imported guideline, because a candidate applying another country's rule will state it confidently and then see exactly where it diverges from Australian practice. Try it with free sample questions at iatroX. For the general problem of applying another country's rules, see cross-country guideline contamination.
Frequently asked questions
What is the format of the RACGP AKT? One hundred and fifty questions using two formats: single best answer items with five options, and extended matching items with a longer option list. The pass mark is set by modified Angoff and varies by cycle, and there is no negative marking.
Why do internationally trained doctors struggle with the AKT? Because it tests Australian general practice rather than general medicine. Therapeutic guidance, prescribing and subsidy, preventive programmes, cardiovascular risk assessment and referral pathways are all national, and excellent medicine from elsewhere produces plausible wrong answers.
What should I revise that I probably have not? The organisational, ethical and health-system content. It is finite, learnable and examinable, and it is reliably postponed because it does not feel like medicine. Space it from the beginning rather than reading it at the end.
How should I approach extended matching questions? Work the case fully and decide on the discriminating feature before you look at the option list. The list is deliberately populated with closely related choices, so scanning first and choosing what feels familiar is the behaviour the format is designed to catch.