AMC recalls: what AMC MCQ recalled questions are, which forms are reliable, and how to use them in preparing for the AMC exam

AMC recalls are questions that candidates reconstruct from memory after sitting the AMC MCQ. They exist because the Australian Medical Council does not release past papers, sample banks beyond a small free application, answer keys or marked scripts. For international medical graduates, a recall is therefore the only available record of what the AMC exam asks. The relevant questions for a candidate are not whether recalls should be used, but which forms of recall material are reliable and how they fit into a preparation plan.

Why AMC recalls exist

Most medical licensing examinations provide official practice material. The General Medical Council publishes PLAB sample questions; the USMLE program releases practice forms; NEET-PG question papers from previous years are widely available. The AMC provides a free application of approximately 210 questions and an examination specification that describes the blueprint without worked items.

The AMC also reuses questions. At least half of each 150-question paper is drawn from a calibrated pool of previously used items, with the remainder being new questions under trial. This is what gives recalls their value: a concept examined in one sitting is likely to be examined again, with a different patient age, laboratory value or distractor. A candidate who has already encountered the concept is better placed to recognise it.

AMC recalls therefore fill a gap in the official material. They are an informal record maintained by candidates who have sat the AMC exam, for the benefit of those preparing to sit it.

What a recalled question contains

A candidate leaving a Pearson VUE centre after 150 questions in 3.5 hours retains a limited and approximate memory of individual items. A typical recollection runs to the essentials of the stem (a 52-year-old man with two months of painless rectal bleeding and no family history), the question asked (the most appropriate next investigation), and two or three of the five options, such as colonoscopy and immunochemical faecal occult blood testing.

That recollection is a recall. The stem is approximate, the numbers are rounded, and some options are missing. What survives intact is the teaching point: symptomatic rectal bleeding at 52 requires colonoscopy rather than a screening test. Recalls are, in effect, a record of teaching points carried in partially remembered scenarios.

This also explains the AMC's published warning that reconstructed question sets in circulation contain errors in both stems and answers. The warning is accurate, and it bears on the central issue: the wording of a recall was never going to be exact, so its usefulness depends entirely on whether the compiler identified the teaching point correctly and verified the answer against a current Australian guideline.

The three forms of AMC recall material

Unverified recall files. These are the documents that circulate through messaging groups, typically several hundred questions with an answer letter and no explanation, and they account for most results for "AMC recalls free download". Their stems are first-hand, which is their value. Their answers are whatever the contributor selected, unchecked by anyone, which is their risk. An incorrect answer learned with confidence from such a file is a mark lost in the examination.

Commercial recall banks. Some coaching providers and subscription services package the same raw material, occasionally with brief explanations, and present it as their principal product. The explanations are frequently written by candidates rather than clinicians, and the currency of the underlying guidelines is seldom stated. Where a product's main claim is possession of "the recalls", the appropriate questions are who wrote the answers and when they were last checked.

Rebuilt, explained sets. In this approach, the topics reported after a sitting are collected, the teaching point underlying each is identified, and a new question is written to examine that point, with an explanation and a guideline reference. This is the method used by AMC Recalls. Each monthly set comprises 100 original questions, each with an answer, an explanation and a brief revision note, and is provided free of charge. Eight sets covering February to September 2026 are available. No question is presented as an AMC item; each is written to teach what the AMC exam was examining.

The first two forms indicate what was asked. Only the third reliably indicates the correct answer and the reasoning behind it.

What the rules permit

Discussing one's recollection of an examination is an established part of medical education and is not prohibited. The AMC's candidate conditions prohibit removing examination content from the test centre, recording the screen, and reproducing or distributing the AMC's questions as such. Breaches of those conditions have resulted in loss of candidacy. None of them describes what a recall is.

AMC Recalls operates within those limits by design. The team has no access to the AMC question bank and does not reproduce examination papers. It identifies the teaching point in the topics candidates discuss and writes a new question on that point. No question on the site is claimed to have appeared in any sitting. The full position is set out in the disclaimer.

What recalls provide that textbooks do not

Exposure to Australian management. For candidates trained in India, Pakistan or Bangladesh, this is the principal benefit. Their training produces competent clinicians whose default management differs from Australian practice in ways that the AMC exam tests directly: a fluoroquinolone where Therapeutic Guidelines specify amoxicillin, bowel cancer screening from 50 rather than 45, BCG vaccination at birth where Australia does not routinely give it. Such gaps are not revealed by further reading of general medical texts. A single recall set exposes most of them, typically within the first 50 questions.

Relative weight within the blueprint. The AMC blueprint allocates 12.5 percent of the paper to population health and ethics. Recalls indicate that, within that allocation, screening programs, the immunisation schedule, and mandatory reporting account for most of the questions. The AMC Part 1 high-yield topics guide was compiled from this kind of sitting-by-sitting analysis.

The construction of an AMC question. The AMC asks for the most appropriate next step considerably more often than for a diagnosis. Its distractors are commonly treatments that are standard in South Asia and incorrect in Australia. Its stems are short. Familiarity with this construction, acquired through several hundred recall-style questions, is particularly valuable in a forward-only, computer-adaptive format in which a question cannot be revisited. The AMC CAT MCQ format guide describes the adaptive mechanism in detail.

Incorporating recall sets into a preparation plan

Recall sets serve different purposes at different stages of preparation, and the sequence matters.

Initial assessment. One complete set, attempted untimed before formal revision begins. The

• score is of little importance; the distribution of errors indicates whether the candidate's weakness lies in clinical knowledge, Australian management or examination technique.

During systematic revision. Recall questions for each system, attempted in the week that

• system is revised, with the explanations consulted. The question and the guideline reinforce each other.

An error record. Each incorrect answer recorded as a concept rather than as a question

• ("migraine with aura: combined oral contraceptive contraindicated"), reviewed weekly. Within six weeks the record is an accurate map of remaining weaknesses.

Final month. Recall sets in timed blocks of 30 questions at approximately 84 seconds each,

• with a decision made on each question before moving on. This reproduces the conditions of the AMC exam, which does not permit return to earlier items.

Final fortnight. The two most recent sittings, which reflect current guidelines and recently

• emphasised topics. For a November sitting these are the September 2026 and August 2026 sets.

The how to pass AMC Part 1 on your first try guide places these stages within a 12 to 16 week schedule alongside a full question bank and three timed 150-question mock examinations.

Limitations of AMC recalls

Recalls are incomplete. Reports from a single sitting cover perhaps a third of the paper, and the questions no candidate remembered remain in the pool. They date: bowel cancer screening was extended to age 45 in July 2024 and the adult pneumococcal schedule changed in July 2026, so a 2022 recall carrying the earlier position teaches an incorrect answer. Unverified recalls also carry the errors of the person who recorded them.

Each of these limitations has the same remedy: recent sets, explained sets, and use alongside a structured, blueprint-weighted question bank rather than in place of one. The AMC Part 1 question bank review compares the principal options, and AMC Academic provides 20 free questions from its 2,800-question bank together with month-specific mock examinations built from the same reported topics.

Recommended starting point

Begin with the most recent set on AMC Recalls: the first 50 questions, untimed, with every explanation read, including those for questions answered correctly. Then consult the AMC exam dates guide, select a sitting, and plan backwards from it. The AMC exam pass rate is approximately 47 percent, and most unsuccessful candidates have adequate clinical knowledge. What they lack is familiarity with what the AMC exam asks, and that is the information recalls provide in advance.