What is a good GAMSAT score?

The average overall GAMSAT score is 56 to 58. The last published peer-reviewed cohort had a mean of 57.45 Peer-reviewed, and the overall mean sat between 55.7 and 57.6 across every year from 2005 to 2014 — a remarkably stable figure, which is what makes any modelling of the curve defensible at all.

That stability is by design. ACER equates each sitting so that scores from different papers are directly comparable ACER-official. A harder paper does not produce lower scores; it produces the same scores from a harder set of questions.

Section averages are not the same as the overall average. Written Communication runs about five points higher than the other two sections, so a 60 in Section 2 is an ordinary result while a 60 in Section 1 is comfortably above average. Percentiles differ sharply by section, and applying the overall curve to a section score is a real error.

It depends almost entirely on the country: published cutoffs run 50–59 in the UK, 53–58 in Ireland and 56–74 in Australia. Those are not different standards of difficulty applied to the same pool — they are different selection systems. The Australian figures come almost entirely from one school with unusually wide sub-quotas.

Two warnings about the numbers you will see quoted elsewhere University-published. First, an interviewee average and an admitted medianare not cutoffs — roughly half of the relevant cohort scored below each. UQ’s widely-quoted 65.33 is a median, and its admitted range ran from 51.66 to 81.00. Treating it as a bar overstates it by about 14 points.

Second, a minimum to apply is not a competitive score. Most Australian schools set 50 overall and 50 in every section; UWA sets 55. Meeting it makes your application valid, not competitive. The cutoffs page separates the four kinds of number.

The same three section scores produce different overall scores at different universities, because the double weighting of Section 3 is at each school’s discretion Peer-reviewed.

ACER’s published overall is (S1 + S2 + 2 × S3) ÷ 4. Melbourne, Queensland and both Notre Dame campuses use a plain average of the three sections instead. Sydney uses no overall score at all and ranks on section scores. Wollongong treats GAMSAT as a hurdle and never ranks it.

Three candidates with identical unweighted averages of 62.67 can hold ACER-weighted overalls of 64.5, 61.5 and 59.5 — a five-point spread produced entirely by where their marks sat. The weighted versus unweighted page works that example through in full, and the score calculator gives you both figures.

A 61 is competitive in the UK and Ireland, and marginal in Australia without a strong GPA. It sits somewhere around the 60th to 70th percentile depending on the sitting Modelled. It clears everypublished UK cutoff, the highest of which is Pears Cumbria’s 59, and every published Irish cutoff — though at UCD and RCSI, both of which settled at 58 in 2025, random selection was applied at the cutoff, so matching it did not guarantee an offer.

In Australia it is above the minimum everywhere but below every published interviewee average, so it needs a GPA doing real work alongside it. At Wollongong, where GAMSAT is a hurdle only, a 61 is worth exactly as much as a 75.

An overall 70 puts you in roughly the top 2–7% of candidates, so about one in twenty reaches it. Modelled It clears every published cutoff in the UK and Ireland and all but one Australian sub-quota.

The section arithmetic matters more than the headline. Because Section 3 has by far the widest spread of the three, a 70 there is only about the 87th to 92nd percentile, while a 70 in Section 1 is near the ceiling of that section’s distribution. Under ACER’s weighting, Section 3 is also half the overall score — which is why an overall 70 is usually built on Section 3 rather than on Section 2.

This is the part a course vendor cannot comfortably tell you: for a large share of applicants, GAMSAT is not the binding constraint.

• GPA carries equal or greater weight at most Australian schools.ANU, Deakin, Griffith, Macquarie and UQ all shortlist on a 50:50 split. Melbourne’s final offer stage is 50% interview, 25% GPA, 25% GAMSAT. At Macquarie, GAMSAT drops out of the offer calculation entirely.
• Casper and interviews can outweigh both. Notre Dame runs 30:30:30:10 across GAMSAT, GPA, Casper and bonus points. At Wollongong the interview rank is 50% Casper and 50% admissions bonuses, with neither GPA nor GAMSAT ranked at all.
• Rural, Indigenous and contextual pathways move the bar substantially. Flinders’ cutoff drops from 74 to 69 for rural applicants and to 56 on the humanitarian visa sub-quota. Pears Cumbria drops from 59 to 53 contextually.
• Several pathways waive GAMSAT entirely.Deakin’s Rural Training Stream Tier 1, Melbourne’s Rural Pathway and North Western pathway, Griffith’s First Peoples Health Pathway, Flinders’ Indigenous Entry Stream and Sydney’s Indigenous Facilitated Entry from 2027 all admit without a GAMSAT score.

A 65 with a 6.4 GPA and a 65 with a 5.5 GPA are not the same application. Before deciding your score is the problem, check which half of your file is actually holding you back.

Four options, in the order most people should consider them.

1. Widen the list before you widen the study plan. Published UK cutoffs run 50–59 and Irish cutoffs 53–58. A score that is marginal in Australia can be comfortably above the line at Limerick, Swansea, ScotGEM or Plymouth. Several UK courses do not even require a degree.
2. Work out whether GPA is the real constraint. If your GAMSAT is at the interviewee average and your GPA is below 5.5, another four points of GAMSAT will not fix the file. Schools weighting GPA at 50% or more are the ones to reconsider.
3. Check what your existing score is still valid for.Results last four years in Australia and two in the UK and Ireland, with Ulster a named exception at four. You may have more cycles available on your current score than you think — see how results and validity work.
4. Resit — with realistic expectations. The only peer-reviewed figure available says repeat candidates differ by about four points, with little evidence of further gains on third or later sittings. The resit page sets out that evidence and its limits.

About four points, on the only peer-reviewed evidence that exists — and with little evidence of an upward trend across further sittings. Peer-reviewedMercer et al. (2015) analysed ten years of ACER data and found that “differences in repeat candidates’ scores between the first and a subsequent attempt is relatively small (about 4 points) with little evidence of an upward trend with further sittings.”

Publish that figure with its limits attached, because they are serious:

• Australian data, March sittings only, covering 2005–2014. The test has since shortened from 75 to 62 questions in Section 1 and from 110 to 75 in Section 3.
• The paper does not say whether the four points refers to the overall score or a section score, gives no standard deviation, and reports no distribution — so it cannot tell you what share of candidates improved, stayed flat or scored lower.
• Three of the five authors sit or sat on the GAMSAT Policy Committee and two work for ACER. It is an analysis of the test published by the test’s own administrators, and it contains no limitations section.

Every larger improvement figure in circulation — “3 to 7 points on Section 3”, “8 to 12 points if you scored below 55” — comes from preparation companies without a citation Prep-company estimate. Forum accounts skew upward for an obvious reason: candidates who jump ten points post about it, and candidates who plateau do not.

If you have not sat GAMSAT yet, the useful measurement is not a predicted score — it is which section is costing you marks under a clock. Section 3 is half of ACER’s overall and has the widest spread of the three, so it is where the same hour of work moves the number furthest.

Our free timed practice test reports a raw percentage on our own questions. It deliberately does not convert that into a predicted GAMSAT score, because no public raw-to-scaled conversion exists and a live sitting also contains unscored trial questions. Any tool that shows you a predicted GAMSAT score is inventing it.