Free tool · Raw marks or a historical score

TMUA score converter

Enter your past-paper marks or an existing historical TMUA score. See the historical score, its Forge estimate on the new scale, the selected year’s percentile and a comparison range. Raw-mark tables cover 2017–2023; 2016 uses the 2017 table as an estimate.

What do you have?
Paper year
of 20
of 20

Your raw mark is the number of questions you got right on that paper. Each paper has 20.

2023 score to the new scale

Historical score 6.2. New-scale Forge estimate 5.2, comparison range 4.4 to 5.9. Approximate 2023 cohort percentile 65.8.

Paper 1 score
5.8
from 10 of 20
Paper 2 score
6.1
from 10 of 20
Historical overall score
6.2
New-scale estimate · Oct 2025
5.2
Comparison range
4.4–5.9

2023 cohort percentile

65.8/ 100

From Cambridge’s published 2023 frequencies (5,627 candidates), using the midpoint of tied scores. This compares ranks across different cohorts, not the same candidates.

October 2024
5.1
October 2025
5.2

Same historical percentile, compared with each October cohort. Both are approximate.

The 2023 overall table gives 6.2 for 20 out of 40. It isn't the average of the two paper scores.

The new-scale score is a Forge percentile comparison, not an official UAT-UK conversion or an exam prediction. The range allows for tied scores, chart precision, grouped bars and the two October cohorts.

How the estimate and range work

We locate your score in the selected year’s overall distribution. The 2016–2022 curves are digitised from Cambridge’s original charts; 2023 uses published counts. Tied scores use their midrank, with linear interpolation between reported scores. The selected year describes its candidates as well as its paper: matching percentiles does not establish equal mathematical ability.

The October 2024/2025 comparisons use UAT-UK’s published histograms and the exact 25th, 50th, 75th and 90th percentile anchors from its technical reports. The latest October cohort supplies the main estimate. Chart bars are grouped in half-point steps, so a one-decimal result is indicative, especially at the extremes.

The comparison range spans both October curves and historical ties. It allows ±1 pixel when reading raster bar heights (±0.12 PDF points for the 2019 vector chart) and ±0.5 score points for target-chart grouping. These are conservative sensitivity choices, not official error bounds, confidence intervals or a forecast of your next sitting. It does not include practice-day variation. January shares the scale but has a different distribution; no 2026 conversion is known in advance. Existing marked results keep their recorded scoring model.

Historical table published by Cambridge Assessment Admissions Testing.

Show the full 2023 table
Paper scores, 2023 table
Raw markPaper 1Paper 2
01.01.0
11.01.0
21.01.0
31.01.2
41.92.2
52.73.0
63.43.7
74.04.4
84.65.0
95.25.6
105.8 (your mark)6.1 (your mark)
116.36.5
126.66.7
136.97.0
147.17.2
157.37.4
167.67.7
177.98.0
188.38.4
199.09.0
209.09.0
Overall score, 2023 table (total out of 40)
TotalOverallTotalOverall
01.0216.5
11.0226.6
21.0236.7
31.0246.8
41.0256.9
51.0267.0
61.5277.1
71.9287.2
82.4297.3
92.8307.4
103.2317.6
113.5327.7
123.9337.8
134.2348.0
144.5358.2
154.8368.4
165.1378.6
175.4389.0
185.7399.0
196.0409.0
206.2 (your total)

How to use it

From a past paper to the new scale

  1. 1

    Enter what you have

    Choose the year the test was sat. Enter both raw marks, or choose a TMUA score if you already know your historical overall result.

  2. 2

    Check the historical score

    Raw marks use the selected year's official tables. The overall score comes from your total out of 40, from its own table.

  3. 3

    See the new-scale estimate

    Compare the historical score with the new-scale Forge estimate, year-specific percentile and range. Use the method disclosure to see the sources and assumptions.

UAT-UK's archive has the 2016 papers, worked answers and answer key, but no score conversion table was published for that year. UAT-UK TMUA preparation materials(opens in a new tab)

Scores since 2024

Why the conversion is an estimate

Historical and current scores use different scales. Forge compares percentiles to estimate an equivalent score; different cohorts and interpolation limit that comparison.

What is a good score on the current scale?
  • The computer-based TMUA, sat since October 2024, reports one overall score from 1.0 to 9.0, to one decimal place. Papers 1 and 2 are equated and scaled together, so there are no separate paper scores.

    UAT-UK test scores(opens in a new tab)
  • The scale is set from each year's October candidates: the median is fixed at 4.5 and the 90th percentile at 7.0. January candidates are scored on the same scale.

    UAT-UK TMUA technical report 2025–26(opens in a new tab)
  • UAT-UK doesn't publish raw-mark conversion tables for the computer-based test. Candidates sit different versions of it, which are equated after the testing window closes, so there's no single table to look up.

    UAT-UK test scores(opens in a new tab)

FAQ

Questions about converting scores

Can I enter a TMUA score instead of raw marks?

Yes. Choose ‘A TMUA score’, select the year the test was sat and enter its overall score from 1.0 to 9.0. The year changes the percentile and new-scale comparison. You will see October 2024 and October 2025 estimates and a comparison range. Scores from 2024 onwards are already on the new scale.

Is the new-scale conversion official?

No. The historical raw-mark tables are official; the new-scale comparison is a Forge estimate. It uses each year’s own overall distribution: chart-derived estimates for 2016–2022 and published counts for 2023. It compares that rank with UAT-UK’s October 2024 and 2025 histograms, retaining the exact percentile anchors in its technical reports. The range allows for ties, chart reading, grouping and differences between October cohorts. Equal rank in different cohorts does not establish equal ability, and January has a different distribution on the same scale.

Why does the same raw mark give different scores in different years?

The conversion was set separately for each year's papers and candidates, so the same raw mark can give a different score. For example, 12 out of 20 on Paper 1 converts to 5.8 in 2018 but 6.9 in 2022.

Is the overall score the average of the two paper scores?

No. The overall score comes from a separate table for your total out of 40. In 2019, 12 on Paper 1 and 10 on Paper 2 give paper scores of 6.7 and 4.8, which average 5.75, but the overall table converts the total of 22 to 5.9.

Why is there no 2016 table?

No conversion table was published for the 2016 papers. Raw-mark mode uses the 2017 score table and that year’s distribution, labelled as an estimate. If you have a genuine 2016 overall result, score mode uses the original 2016 distribution instead.

Can I convert raw marks from the 2024 or 2025 test?

No. The computer-based test reports one score from 1.0 to 9.0, and UAT-UK doesn't publish raw-mark tables for it. The tables here cover the paper-based tests up to 2023, and their scores aren't directly comparable with the current scale.

Your next step

Choose your next practice

Free, no account needed

Take the timed diagnostic

20 Forge-original questions under exam timing, calibrated to the 2023 paper's difficulty profile. You get a score and a topic-by-topic breakdown at the end.

Start the free diagnostic

Free account

Save your progress

Create a free account for 41 free lessons and 4 methods in Learn, with progress kept on this device, and free sittings of released past papers.

Create a free account

Premium

Practise at the harder level

561 Forge originals calibrated for the harder computer-based papers sat since 2024, with worked solutions, unlimited mocks, all 104 lessons and 18 methods.

£49 for 12 months or £20 for 30 days. One payment, no automatic renewal.

See premium