HSC scaling explained: how UAC scales every course, the 2025 scaling report table (scaled mean, 90th and 50th percentile by subject), the aggregate-to-ATAR table, the myths, what it means for subject choice, and a printable scaling table

HSC scaling is the part of the ATAR that nobody at school explains properly, so families work from rumours: "Physics scales up", "Standard Maths scales down", "do Extension 2 for the scaling". The Universities Admissions Centre publishes a full report on how it scaled each year's HSC, with the scaled mean and percentiles of every course, and the rumours are mostly wrong in the same way: they confuse the strength of a course's candidature with what a given child will score in it. This page is the 2025 report put into plain tables (the one for the 2026 HSC comes out early in 2027 and will look much the same), with the aggregate-to-ATAR conversion, UAC's own worked examples, and the only rule of subject choice that the numbers support. Also a printable of the table and the checklist.

Download the 2025 HSC scaling table and subject-choice checklist (PDF) (35 courses with the scaled mean and the scaled mark at the 90th, 50th and 25th percentiles, the aggregate-to-ATAR table, the ATAR percentages and the checklist; two pages, free).

What scaling is, in one paragraph

The ATAR is a rank, not a mark: 90.00 means the student is in the top 10 percent of the age cohort had everyone sat for an ATAR. To rank students who took different courses, UAC's Technical Committee on Scaling converts every course's raw marks onto one scale. For each course it looks at how its students performed across all their other courses, and sets the course's scaled mean and standard deviation to match that. A course whose students are strong everywhere gets a high scaled mean; a course whose students are weaker everywhere gets a low one. Then each student's scaled mark in the course is set by their position in the course's raw-mark distribution. Two things, and only two, decide a scaled mark: the student's rank in the course and the strength of the course's candidature. The content, the "hardness" of the syllabus and the reputation of the subject are not inputs.

The scaled marks are out of 50 per unit. The aggregate is the sum of the best ten units: two units of English (compulsory) plus the best eight other units, so it is out of 500. The aggregate is placed on a percentile scale and converted to the ATAR with a model that depends only on the size of the age cohort; since 2025 the aggregate is kept to two decimal places, which is why ties became rare (the most students sharing one aggregate in 2025 was 10, against 30 or more in 2024).

The 2025 scaling table: scaled mean and percentiles by course

From Table A3 of UAC's Report on the Scaling of the 2025 NSW HSC. The scaled mean is the average scaled mark of the course's students; P90, P50 and P25 are the scaled marks of the students at the 90th, 50th and 25th percentiles of the course. "HSC mark at P50" is the HSC mark (the one on the certificate, per unit) of the middle student, for comparison. All marks are per unit out of 50.

Three things to read off the table before anyone says "scales well":

1. The HSC mark on the certificate is not the scaled mark. The middle student in Visual Arts has an HSC mark of 41.5 and a scaled mark of 21.1; the middle student in Chemistry has 38.0 and 33.9. HSC marks are aligned to performance bands and sit between 70 and 80 (out of 100) for the median of most courses; scaled marks are positions. This is why the ATAR looks low next to the marks: the median ATAR in 2025 was 70.75 and the median HSC mark in most courses was higher.
2. The spread inside a course is larger than the gap between courses. The 90th-percentile student in Business Studies (39.2) outscores the median student in Chemistry (33.9) by five marks per unit. The 25th-percentile student in Chemistry (25.8) is below the median student in Business Studies.
3. English Standard and English Advanced are scaled as one group. NESA puts them on a common scale with shared examination items, so a given level of English achievement gives the same scaled mark in either course. The Standard scaled mean is lower because its candidature is weaker, not because Standard is penalised: a Standard student who would have been in the middle of Advanced gets the Advanced-middle scaled mark.

From aggregate to ATAR in 2025

Table 3.2 of the report. The aggregate is the sum of the ten best unit scaled marks, out of 500.

So an ATAR of 90 needed an aggregate of about 365, an average scaled mark of about 36.5 per unit over the best ten; an ATAR of 80 needed about 315. The share of ATAR-eligible students at each level, 2021 to 2025 (Table 4.4): 99.00 and above, 1.7 to 1.8 percent; 95.00 and above, 8.6 to 8.9 percent; 90.00 and above, 17.1 to 17.8 percent; 80.00 and above, 34.2 to 35.5 percent; 70.00 and above, 50.7 to 52.6 percent. The median ATAR was 70.75 in 2025 (71.60 for female students, 69.80 for male), and 53 students received 99.95. An ATAR of 99.00 is not the top 1 percent of the ATAR cohort (1.7 percent got it); it is the top 1 percent of the whole age group had everyone qualified.

Of the 77,693 students who completed at least one ATAR course in 2025, 77.8 percent received an ATAR; 8,771 completed a full HSC that did not qualify for one, usually because too many units were Board Endorsed courses.

UAC's own examples: position beats scaled mean

The report answers the two questions that fill its enquiry centre every December, and its examples are better than any tutoring company's.

Same HSC marks, different ATARs (Table 6.1). Liam and Kellie each sit at the 50th percentile of all their courses with an average HSC mark of exactly 40.0 per unit. Liam's ATAR is 60.00 and Kellie's is 80.00, because Kellie's courses have higher scaled means. Same marks, same positions, different candidatures.

Same HSC mark, higher scaled mark in the "worse-scaling" course (Table 6.6). A Food Technology student with an HSC mark of 48.0 per unit is at the 99th percentile of the course and receives a scaled mark of 43.3. A Korean Beginners student with the same 48.0 is at the 90th percentile and receives 42.8. Food Technology's scaled mean is 18.9 and Korean Beginners' is 27.7, nine marks higher, and the Food Technology student still wins, because position in the course is the first input and the candidature only the second.

Same position, same mean, different spread (Table 6.7). Students at the 90th percentile of Chinese Extension (scaled mean 37.3) and Music 2 (scaled mean 34.1) both have HSC marks of 48.0; the Music 2 student's scaled mark is 44.9 and the Chinese Extension student's 44.4, because Music 2's candidature is more spread out (standard deviation 8.5 against 5.4), so its top students sit further above its mean.

Why one course counted and a higher-marked one did not. Each student is told which units made their aggregate. A course with a lower HSC mark can contribute ahead of one with a higher mark when the student's position in the lower-marked course, or that course's scaled mean, gives it the higher scaled mark. The report's answer is that "a student's position in their course and the scaled means and standard deviations of their courses are all important", and that one HSC mark corresponds to a range of scaled marks, so no conversion table exists.

What the numbers say about choosing subjects

• Choose the course the child will rank highly in. The 99th-percentile Food Technology student beat the 90th-percentile Korean student. A child in the top 10 percent of Business Studies (39.2) is worth more to the aggregate than a child in the bottom quarter of Chemistry (25.8), by thirteen marks per unit.
• Where the child would rank the same, the higher scaled mean wins. This is the only legitimate "scaling" choice, and it is usually a choice between two courses the child is equally good at: Economics over Business Studies for a child who reads well and likes argument; Chemistry over Biology for a child who is comfortable with maths. The gap at the median was 9.6 marks per unit between Economics and Business Studies and 7.2 between Chemistry and Biology.
• Mathematics: take the highest level the child can rank in the top half of. The scaled mean rises from 23.3 (Standard 2) to 31.9 (Advanced) to 39.7 (Extension 1) to 43.4 (Extension 2), but a child in the bottom quarter of Advanced (25.9) is only slightly ahead of the median Standard 2 student (23.2), and well behind the top 10 percent of Standard 2 (37.3). The Advanced course is the right choice for a child who will sit in its top half; for a child who would be in its bottom quarter it is close to a wash, with a year of stress added.
• English counts whatever happens. Two units of English are in every aggregate; it is the one course where improving the rank always improves the ATAR, which is why English tutoring has the most reliable return of any subject in the HSC.
• Extension courses are worth it for the students who belong in them. The 25th-percentile student in Extension 2 Mathematics still scored 41.5 per unit, above the 90th percentile of eleven courses in the table. The same is not true of a child who scrapes into Extension 1 and ranks last.
• Ignore the subject's reputation and the school's advice about "scaling" unless it is about rank. The report does not scale Physics up or Standard Maths down; it reads the students. In 2025 Drama's median HSC mark was 41.0 and its scaled median 23.4; Chemistry's were 38.0 and 33.9. Neither number is about how hard the course is.

How we use this with HSC students

We teach Year 11 and 12 students in Sydney and Melbourne 1-on-1 online in the after-school hours, and the scaling table is on the desk at the subject-choice meeting in Year 10 for one purpose: to find the two or three courses the child can rank in the top quarter of, and to make English one of them. Then the work is rank, not scaling: past papers marked to the NESA guidelines, the trial as a dress rehearsal, and the mark that moves is the raw mark that sets the position. The first 30-minute class is free; send the subject list and the last report on WhatsApp or the contact page. The papers are in HSC past papers, the reference sheet in HSC maths formula sheet, the Victorian version in VCE scaling explained, and the city pages are Sydney and Melbourne.

Download the 2025 HSC scaling table and subject-choice checklist (PDF)

Source: Universities Admissions Centre, Report on the Scaling of the 2025 NSW Higher School Certificate (Technical Committee on Scaling, published early 2026), Tables 3.1, 3.2, 4.4, 4.5, 6.1, 6.5 to 6.7 and A3, read on 25 October 2026. The 2026 report will replace these figures when UAC publishes it; the pattern between courses changes little from year to year.

Questions parents ask

See how we teach this, 1-on-1 online →