The AMC Multiple Choice Question examination is a Computer Adaptive Test. Each candidate answers a unique sequence of 150 questions built in real time from their own responses, delivered in one 3.5 hour sitting.
The adaptive mechanic
The first question is drawn at random from the item pool. Every question after that depends on the one before it. A correct answer moves the next question to a higher difficulty level, an incorrect answer moves it lower. After each response the system recalculates the candidate's ability estimate from everything answered so far, and that estimate sharpens through the session because each new question targets it directly.
This branching sits on top of item response theory, the statistical framework behind adaptive testing generally. Every question in a pool carries a measured difficulty. The system chooses each next item to extract the most information at a candidate's current ability level, narrowing that estimate with every response. This is what lets an adaptive test reach the same precision as a much longer fixed-form test, using fewer items.
The calibration pool
At least half of a candidate's 150 questions come from a pool of previously calibrated items, drawn from data on how earlier candidates performed on them. The rest are new, not previously used in an AMC exam. Responses to these are inspected and calibrated before the items are used for scoring.
Blueprint and equating
The 150 questions are blueprinted across six patient groups at fixed weights: adult health medicine at 30%, adult health surgery at 20%, and women's health, child health, mental health, and population health and ethics at 12.5% each. Each question is also classified by the clinical task it tests, whether that is gathering data, interpreting and synthesising it, or managing the patient. Every CAT sitting is automatically equated against every other test drawn from the AMC's pools, so the pass point holds regardless of which specific items a candidate was given.
No review
A question must be answered before the next one is generated, and a submitted answer cannot be revisited or changed. The AMC introduced this rule for exam security, bringing its CAT MCQ in line with other adaptive licensing exams internationally.
How the score is built
The ability estimate weighs the difficulty of the questions actually administered. Two candidates who answer the same number correctly can still receive different results, because one may have been routed harder questions than the other. The result sits on a scale of 0 to 500, with a pass represented at 250. A candidate who does not complete all 150 items risks a result of Fail, insufficient data, because the system has not gathered enough responses to settle on a reliable estimate.
From 2026 the AMC is raising the pass standard slightly, still reported on the same 0 to 500 scale with the pass mark described as 250. The AMC frames the change as routine benchmarking against the level expected of graduating Australian medical students.
What this means for preparation
A test that reroutes itself after every answer rewards steady, well-founded reasoning sustained across the full breadth of the syllabus. IMGIIAMC is built by clinicians around exactly that discipline, one question at a time.