GAMSAT practice questions

The pacing that falls out of those numbers is the first thing worth internalising: about 97 seconds per question in Section 1 and 120 seconds in Section 3, including the time spent reading the stimulus they hang off. Practising questions untimed teaches you the content but not the constraint, and the constraint is most of what makes GAMSAT hard.

ACER no longer publishes a reading-time breakdown. Earlier booklets (2020–2022) itemised it; the 2026 edition does not.

• It belongs to a shared-stimulus unit. Questions arrive in groups of 1 to 7 attached to one stimulus, most commonly 2–5. A page of standalone, unrelated items is a quiz, not a GAMSAT simulation — it removes the cost of reading a stimulus, which is where the time goes.
• No calculator. Every arithmetic step has to be doable by hand inside the time. A question that needs a calculator has been written for a different exam.
• No formula sheet and no periodic table. Any constant, value or relationship a question needs must be supplied in its own stimulus. This is the property most third-party question banks get wrong, and it changes the skill being tested: GAMSAT rewards extracting what you need from what is in front of you, not recalling it.
• Every question is worth the same. ACER states: "All questions have the same value." Marks are NOT weighted by question difficulty.A practice set that weights “hard” questions more heavily is scoring you on a system ACER does not use.
• No penalty for a wrong answer. ACER deducts nothing, so leaving a question blank is strictly worse than guessing. Practice material that punishes guesses trains the wrong instinct.

There are no GAMSAT past papers in the sense that exists for a school exam. ACER does not release used papers. What it does sell is a set of preparation materials, and it is explicit that these are “the only official GAMSAT preparation materials”. You do not need to be registered to buy them.

Two things follow from that table. First, if you have registered you already own Practice Test A — a full-length equivalent of the two test-centre sections, and the best single piece of material in existence for calibrating yourself. Second, the free preparation videos and the platform demo cost nothing and are the only way to see the real test application before test day.

Which brings us to the PDFs. The “GAMSAT past papers” circulating on file hosts, forums and messaging groups are one of two things, and both are bad buys at any price, including free:

• Redistributed ACER material. Copied and shared in direct breach of the licence above. ACER treats using test content for any purpose other than your own sitting as misconduct, and it lists publishing questions online and passing them to third parties by name. Sanctions run to cancelled registrations without refund, withheld results, and disqualification from future sittings.
• Reconstructions from memory.Someone’s recollection of questions they saw, typed up. Unreliable by construction, frequently years out of date, and often built on a format ACER has since changed. Plenty of material still in circulation describes a single-day exam with itemised per-section reading time; neither matches the exam ACER documents today.

We take a clean-room position on this deliberately. We do not buy ACER’s materials, because accepting that licence would compromise the independence that makes our questions defensible. We write from ACER’s free published specification instead. That is set out in full on what we are and what we are not.

ACER runs plagiarism checks on every Section 2 response as part of the marking process. The checks use AI, they run against previously submitted GAMSAT responses, and ACER names the reviewer it uses. Then it says this, verbatim:

“Essays that borrow text or statements from pre-prepared essays (often provided by preparation providers) could be flagged as plagiarised even if you have not seen the original source. In these cases you could be plagiarizing without even knowing it, and may be subject to misconduct clauses.”

Read that as the operational warning it is. The risk is not only to a provider who recycled ACER’s prompts — it lands on you, in your own marking, for phrasing you may have absorbed without ever seeing where it came from. ACER adds that memorised essays generally score poorly regardless, and recommends avoiding the approach entirely.

The same logic extends to the multiple-choice sections. ACER states that candidates exposed to real GAMSAT questions at a coaching course risk having their scores withheld, and lists “intended or unintended exposure to GAMSAT content” among its triggers for withholding a result. Unintended is doing a lot of work in that sentence.

Original questions written to the published specification carry none of this exposure. That is not a marketing line; it is the whole reason our material is written from scratch rather than harvested.

A percentage on a practice set is a real measurement of your performance on that set. It is not a GAMSAT score, and it cannot be converted into one.

A live GAMSAT sitting includes unscored trial questions. Neither ACER nor any third party can reproduce an exact scaled score from raw performance.ACER scales raw marks onto a 0–100 scale calibrated against everyone who sat that particular form, calculates the overall score before rounding the section scores, and states outright that this makes the process impossible for a candidate to replicate.

So use practice questions for the two things they genuinely measure: whether you can finish in time, and which topics collapse under pressure. Anything advertising a “predicted GAMSAT score” from a practice set is inventing the number. More on how the real scoring works in GAMSAT results explained.

Start with the timed test, then use the section pages to work out what to fix. The syllabus page is the map of what Section 3 can actually ask about, which is narrower than most candidates fear.