FOI data analysis · 2020–2024 entry · Cambridge Engineering & Natural Sciences
Inside Cambridge Admissions: What the FOI Data Shows
Cambridge does not publish ESAT score data linked to applicant outcomes. Unlike Imperial College London — which provides a public outcome dashboard — Cambridge's admissions statistics are almost entirely opaque. This post analyses three Freedom of Information responses obtained from the University of Cambridge covering Engineering (H100) and Natural Sciences (BCF0) — the only public source that links individual ESAT/ENGAA scores to offer status by domicile.
The data covers 12,063 individual ESAT/ENGAA applicant records (2020–2024) and 2,215 A-level qualification records (2020 cohort), each tagged with domicile (Home, China, International) and offer status. The most striking finding is what the data reveals about the systematic differences between how applicants from different domiciles are selected — not in raw score thresholds, but in the score required to achieve the same offer probability.
12,063Individual applicant records (2020–2024)
4,962ESAT records analysed (2023 & 2024 entry)
17.8Median total score of offer holders
10.1Median total score of non-offer holders
What data was obtained and how
Three FOI requests · Cambridge University · 2024–2025
FOI Reference
Requester
Course
Data Provided
Years Covered
FOI-2024-484 (Lim)
Winter Lim
Engineering H100
Individual ESAT/ENGAA section scores by domicile and offer status · 2020 A-level qualification types and predicted grades
2020, 2022, 2023, 2024
FOI-2025-1026 (Wang)
Wang
Natural Sciences BCF0
Individual ESAT section scores by offer status · Aggregate domicile breakdown table (no individual domicile column)
2025 entry
FOI-2025-347 (Stafford)
Stafford
CS G400 & Economics L100
Individual TMUA scores (Mathematical Thinking, Mathematical Reasoning, Overall) by offer status
2024 entry
FOI-2025-1028 (Smith)
Smith
Engineering H100
Min/avg/max ESAT scores by college, for all applicants and offer holders separately, across all three Engineering sections
2025 apply year
⚠ Interview data not available: Cambridge explicitly refused to provide interview scores in all FOI responses. The University states that interview processes are "set and managed locally by each of the individual Colleges, each of which is a separate legal entity and public authority." No university-level interview data exists. All analysis in this post is therefore based on pre-interview inputs (ESAT scores, A-level grades, domicile).
Understanding the ESAT sections at Cambridge
Cambridge requires three sections per applicant — the combination varies by course
ESAT section scores (scaled 1.0–9.0, same as Imperial)
Data sources used in this report: FOI-2024-484 (Lim) Engineering H100 individual data · FOI-2025-1026 (Wang) Natural Sciences BCF0 summary data · FOI-2025-347 (Stafford) CS/Economics TMUA data · FOI-2025-1028 (Smith) Engineering ESAT by college · Cambridge Admissions Statistics (published) · UAT-UK ESAT Technical Reports.
Engineering H100 · 2023 & 2024 entry · n = 4,962 · S1 = first ESAT section score
ESAT Scores & Offer Rates
The Lim FOI data contains individual section scores for 4,962 Engineering applicants across the 2023 and 2024 cycles — uniquely allowing offer rates to be computed at each score band. The first ESAT section (S1, corresponding to Maths 1) is the strongest single predictor of offer outcomes. The relationship is steep: below S1=4, offer rates are in single digits for all domiciles. Above S1=7, Home applicants receive offers at nearly 90%.
Offer rate by S1 score band — Home vs China vs International (2023+2024)
Score band breakdown — all three domiciles
Engineering H100 · 2023 & 2024 combined · Raw offer rates per S1 band
S1 Band
Home — n
Home — offer rate
China — n
China — offer rate
International — n
International — offer rate
1–2
1,285
2.1%
37
0.0%
442
0.5%
3
696
11.1%
68
1.5%
285
1.8%
4
527
23.5%
90
1.1%
252
8.7%
5
224
44.2%
112
7.1%
130
15.4%
6
149
57.7%
137
13.9%
129
28.7%
7
33
81.8%
64
20.3%
40
42.5%
8
13
84.6%
56
35.7%
29
41.4%
9
17
100.0%
104
54.8%
43
74.4%
Key observation: At every single score band, the offer rate for Home applicants exceeds that for Chinese applicants — often by a factor of 5–10×. A Home applicant scoring S1=4 has a 23.5% offer rate. A Chinese applicant scoring the same has a 1.1% offer rate. This is the clearest quantitative signal that domicile profoundly shapes the offer probability at any given test score.
Total score distribution — offer holders vs non-offer holders
Sum of all three ESAT sections (max 27.0) · 2023 & 2024 entry
🎓 Engineering H100 — overall
All domiciles combined · 2023 & 2024 entry · n = 4,962
15.0%
combined offer rate
Offer holders
18.1
mean total /27
median 17.8
25th–75th pct (offers)
14.5–21.8
total score range
interquartile
Non-offer holders
10.7
mean total /27
median 10.1
Mean total ESAT score (out of 27) by domicile — offer holders
Home
16.0
China
23.8
International
20.3
Chinese applicants who receive offers average a combined ESAT score of 23.8 out of 27 — nearly 8 points higher than Home offer holders at 16.0. This is not because Cambridge applies a formal higher threshold for Chinese applicants; it is because the applicant pool for China skews heavily toward the top of the score distribution, while competition for the same fixed number of places is intense.
What score is enough to expect an offer?
S1 score at which offer rate exceeds 50% — by domicile
50% offer rate threshold by domicile →
Home
S1 ≥ 6 crosses 50% offer rate (57.7%). At S1 = 7+, 87.3% of Home applicants received offers. The curve is steep: each additional point above 5 roughly doubles the offer probability. Home applicants with S1 = 9 received offers at 100%.
China
No S1 band crosses 50% for Chinese applicants. Even at S1 = 9, the offer rate is 54.8% — barely above the threshold. At S1 = 7, it is only 20.3%. Achieving what a Home applicant achieves at S1 = 6 requires a Chinese applicant to score 9.0. The gap is structural, not incidental.
International
S1 ≥ 9 crosses 50% (74.4%). The International curve lies between Home and China — significantly below Home at all score levels, but better than China above S1 = 6. At S1 = 7+, International offer rate is 42.5%, compared to 20.3% for China and 81.8% for Home.
The most striking pattern in all three FOI datasets is the systematic gap between the ESAT score needed to receive an offer at Cambridge, depending on domicile. This tab explores that gap in depth — across Engineering and Natural Sciences — and what the Winter Pool data reveals about how Chinese applicants fare after the interview stage.
Average S1 scores — offer holders vs non-offer holders
The 2.72-point gap: Chinese offer holders average an S1 of 7.95, versus 5.23 for Home offer holders — a gap of 2.72 points on a 9-point scale. In percentile terms, this is the difference between approximately the 70th and 95th percentile of ESAT test-takers. Chinese applicants must score at the elite end of the national distribution to achieve the same outcome probability as an above-average Home applicant.
Avg S1 score — offer holders vs non-offer holders by domicile
Natural Sciences: the Winter Pool picture
FOI-2025-1026 (Wang) · Natural Sciences BCF0 · 2025 entry · Domicile summary table only (no individual domicile column in score data)
What is the Winter Pool?
Cambridge's Winter Pool is a redistribution mechanism. If a college interviews an applicant but declines to offer directly, it may place them in the pool. Other colleges with spare places then review pooled applicants and may invite them for a second interview and potentially offer a place. Pool placement ≠ an offer — it is a second chance. The pool conversion rate (how many pooled applicants eventually receive offers) varies significantly by domicile.
Domicile
Applications
Direct offers
Direct offer rate
Pool placement
Pool placement rate
Pool offers
Pool conversion
Total success
UK
1,381
446
32.3%
308
22.3%
~101
32.8%
39.6%
China
722
108
15.0%
151
20.9%
~8
5.3%
16.1%
Singapore
81
32
39.5%
—
—
—
—
—
Hong Kong
67
20
29.9%
—
—
—
—
—
USA
78
8
10.3%
—
—
—
—
—
🔁 The pool conversion disparity
Natural Sciences BCF0 · UK vs China · 2025 entry
UK pool conversion
32.8%
of pooled UK applicants received a pool offer
~101 pool offers
China pool conversion
5.3%
of pooled Chinese applicants received a pool offer
~8 pool offers
Pool placement rate
~21%
both UK and China enter pool at similar rates
comparable placement
The pool placement rate is nearly identical — about 21% for both UK and Chinese applicants — meaning colleges are equally willing to put both groups into the pool rather than outright reject them. But pool conversion is 6× lower for Chinese applicants. Since pooled applicants have already been interviewed by at least one college, the pool conversion gap likely reflects interview-stage factors: communication, cultural expectations set by individual college Fellows, or logistics of second interviews for overseas candidates.
The structural picture: Chinese applicants enter the pool at the same rate as UK applicants (~21%), which means they are clearing the initial application and ESAT bar at a comparable rate once competition intensity is controlled for. But once in the pool — at the interview conversion stage — they succeed at one-sixth the rate. This is the clearest data signal that the interview, not the ESAT, is where the domicile gap is amplified.
2020 cohort only · Engineering H100 · Home A-level applicants · Predicted grades at time of application
A-Level Grades: Floor, Not Differentiator
The Lim FOI contains a separate dataset for the 2020 cohort: qualification type and predicted A-level grades for each applicant, alongside offer status and domicile. With 2,215 records, this allows direct comparison of predicted grade profiles between successful and unsuccessful applicants. The central finding is counterintuitive: A-level grades are almost completely non-discriminating between offer and non-offer holders at Cambridge Engineering. The same top grade profile appears in both columns.
3.46Average A* count offer holders (Home)
2.62Average A* count non-offer holders (Home)
36%Offer rate for 4-A* predicted
57.1%Chinese applicants using non-A-level quals
Offer rate by number of predicted A* grades
Home A-level applicants only · 2020 cohort · n = 1,202
Offer rate by predicted A* count — Home Engineering applicants (2020)
📋 Most common predicted grades — offer vs non-offer
Home A-level applicants · Engineering H100 · 2020 · Top 6 grade strings each group
A*A*A*A* is the most common predicted grade string for both offer holders (113 applicants) and non-offer holders (202 applicants). The non-offer group has proportionally more applicants with 4 A*s than the offer group. This conclusively demonstrates that predicted grades do not drive final selection — they function as a minimum bar. Selection happens at the ESAT and interview stage.
Chinese applicants: qualification type breakdown
2020 cohort · n = 196 Chinese applicants in Engineering H100
Chinese applicant qualification profile →
A-level (57.1%)
112 Chinese applicants held A-level qualifications. Of these, 25.9% (29 applicants) received offers. Chinese A-level applicants tend to sit A-levels at international schools or UK boarding schools and submit the same qualification profile as Home applicants.
Other (42.9%)
84 Chinese applicants were classified as "Other" qualifications — including Gaokao, Chinese high school diplomas, IB, and other international curricula. Remarkably, their offer rate was higher: 36.9% (31 applicants), compared to 25.9% for Chinese A-level applicants. This suggests Cambridge's "Other" pathway for Chinese applicants may select for particularly strong candidates.
Overall (2020)
Combined Chinese offer rate: 30.6% (60/196). This appears higher than the 2023–24 figure of 17.8% — suggesting either 2020 had unusually high intake, or that the qualification data sample is not fully representative of the ESAT score cohorts (different IDs, different years, different selection criteria).
⚠ Limitation: The A-level data is from the 2020 cohort only — the year before ESAT was introduced. In 2020, Cambridge used the ENGAA, not the ESAT. Grade profiles from this cohort may not perfectly represent the current competitive landscape, though the structural relationships (grades as a floor, not a differentiator) are unlikely to have changed significantly.
The Wang FOI data for Natural Sciences allows offer rates to be computed for each ESAT section independently — separating the signal from each section. The result reveals a striking asymmetry: Maths 1 is the primary predictor for both streams, and Physics and Maths 2 dominate Physical NatSci selection. Biology and Chemistry — the "subject" sections for Biological NatSci — show weak, non-linear, and sometimes inverse relationships with offer rate. This has a direct practical implication for applicants.
1,160Biological NatSci applicants (2025)
24.8%Biological NatSci offer rate
1,465Physical NatSci applicants (2025)
26.2%Physical NatSci offer rate
Biological NatSci: Maths 1 rules, Biology and Chemistry do not
Offer rate by section score band — Biological stream · n = 1,160 applicants · 288 offers
⚠ The Biology and Chemistry averages for offer holders are lower than for non-offer holders. This is not a data error — it reflects that high subject-section scores do not compensate for a weak Maths 1, while moderate subject scores paired with a strong Maths 1 do lead to offers.
Biological NatSci — offer rate by section score band
Critical finding — Biology and Chemistry are not reliable predictors: Across the score range 1.0–9.0, the Biology offer rate chart is non-linear and erratic, peaking around 3.0–3.9 at ~80%, then dropping, then partially recovering. Both sections show negative correlation with offer rate above the 5.0 band — meaning applicants who scored very highly in Biology or Chemistry were less likely to receive offers than those who scored moderately.
Physical NatSci: Physics and Maths 2 are decisive
Offer rate by section score band — Physical stream · n = 1,465 applicants · 384 offers
Physics and Maths 2 gaps between offer and non-offer holders are 3.17 and 3.44 points respectively — among the largest section-level gaps observed across any Cambridge or Imperial dataset in this analysis. Virtually no offers are made below Physics = 5 or Maths 2 = 6.
Physical NatSci — offer rate by section score band
What the section data tells applicants →
For both streams
Maths 1 is the single most consistent predictor of offer outcomes across all NatSci streams. If you can only improve one section, Maths 1 gives the greatest expected return on offer probability.
Physical NatSci
Physics and Maths 2 are both decisive — they need to be strong simultaneously. Offer rates near zero below Physics = 5 and Maths 2 = 6. Aiming for 7+ in both gives near-certainty of being competitive.
Biological NatSci
Even for the Biological stream, stronger Biology and Chemistry scores do not meaningfully increase offer probability. Treat Biology and Chemistry as sections to pass competently (aim for 4.0–6.0) rather than to maximise.
Two college-level datasets are available from the FOI responses. The TMUA table covers CS (G400) and Economics (L100) offer holders for the 2024 application cycle, derived from FOI-2025-347 (Stafford). The Natural Sciences table covers all BCF0 applications for the 2025 cycle, derived from FOI-2025-1026 (Wang). Neither dataset includes individual domicile-level breakdowns at the college level — figures are pooled across all domiciles.
Note that Cambridge admissions are college-based, not department-based. The same applicant pool is divided across 29 colleges, each operating its own selection process independently. College-level variation in scores and offer rates therefore reflects both the differing academic cultures of individual colleges and random variation from small sample sizes — particularly for smaller colleges.
Average TMUA scores by college — offer holders, 2024
⚠ Note on Wolfson and Hughes Hall: Both are primarily graduate colleges with very small undergraduate cohorts. Their high averages reflect a tiny and self-selected applicant pool rather than a systematically higher bar. Newnham had no offer holder TMUA data recorded for the 2024 cycle. The TMUA data covers CS (G400) and Economics (L100) only — it does not cover Engineering or Natural Sciences, which use the ESAT instead.
Natural Sciences (BCF0) applications by college, 2025
All colleges · Sorted by total applications (highest → lowest) · <3 = suppressed to protect individual privacy
#
College
Applications
Offers
Offer rate
Winter pool placed
Pool offers
1
Trinity
171
37
21.6%
44
0
2
Emmanuel
155
28
18.1%
33
0
3
St Catharine's
141
33
23.4%
15
0
4
Churchill
128
42
32.8%
28
8
5
St John's
125
29
23.2%
23
10
6
Clare
123
27
22.0%
31
6
7
Fitzwilliam
118
27
22.9%
20
0
8
Pembroke
118
30
25.4%
29
0
9
Jesus
117
27
23.1%
21
0
10
Christ's
112
31
27.7%
17
3
11
Gonville & Caius
109
24
22.0%
22
<3
12
Homerton
105
33
31.4%
25
9
13
Downing
104
21
20.2%
24
<3
14
Queens'
88
26
29.5%
29
<3
15
Magdalene
88
19
21.6%
23
0
16
Selwyn
86
22
25.6%
21
4
17
Newnham
83
20
24.1%
16
6
18
Robinson
81
24
29.6%
9
8
19
Trinity Hall
81
22
27.2%
26
4
20
Peterhouse
77
15
19.5%
15
3
21
Lucy Cavendish
77
25
32.5%
5
13
22
Sidney Sussex
68
19
27.9%
11
10
23
Murray Edwards
66
18
27.3%
7
12
24
Corpus Christi
65
20
30.8%
15
3
25
King's
62
14
22.6%
19
3
26
Girton
79
24
30.4%
16
16
27
Wolfson
6
<3
—
<3
0
28
St Edmund's
6
9
—
<3
5
29
Hughes Hall
5
4
—
<3
3
Pool offers vs pool conversion: Colleges with high pool offer counts relative to their direct offers — notably Lucy Cavendish (13 pool offers against 25 direct), Girton (16 pool offers equal to its direct offer count), and Murray Edwards (12 pool against 18 direct) — actively recruit from the Winter Pool.
⚠ Cross-dataset comparison: The TMUA table covers the 2024 cycle and CS/Economics only. The NatSci table covers the 2025 cycle and Natural Sciences only. The two datasets are not directly comparable.
Engineering H100 · 2025 apply year · FOI-2025-1028 (Smith) · Average scores for all applicants vs offer holders
ESAT Scores by College
This data comes from FOI-2025-1028 (Smith), which provides minimum, average, and maximum ESAT section scores broken down by Cambridge college for the 2025 application year. Scores are given separately for all applicants and for offer holders at each college, across all three Engineering sections: Mathematics 1, Physics, and Mathematics 2.
The chart below shows average scores per college. Use the dropdowns to switch section and sort order. The gap between the two bars at each college indicates how selectively a college converts its applicant pool into offers — a wide gap means only the highest-scoring applicants receive offers.
Avg ESAT score — Mathematics 1 · all applicants vs offer holders · by college
All applicants (avg)
Offer holders (avg)
* Hughes Hall offer holder averages withheld by Cambridge (small cohort, shown as absent bar). St Edmund's shows applicant averages above offer holder averages across all sections — likely reflects pool-in offers attributed to St Edmund's in the FOI data. Wolfson has a compressed score range due to its graduate-entry focus. Scores on 1.0–9.0 scale.
How to read the gap: A wide gap between the blue (all applicants) and green (offer holders) bar means a college received a wide range of applicants but selected only the highest scorers. A narrow gap suggests a more self-selecting applicant pool where most applicants who applied were already competitive. Large-gap colleges are not necessarily "harder" — they may simply attract more speculative applications.
Full data — all 29 colleges, all three sections
Engineering H100 · 2025 apply year · Source: FOI-2025-1028 (Smith) · Averages on 1.0–9.0 scale · — = withheld
College
M1 — all
M1 — offers
Phy — all
Phy — offers
M2 — all
M2 — offers
Christ's
5.00
7.05
4.79
6.60
5.03
6.42
Churchill
4.80
5.98
4.77
5.90
4.83
5.91
Clare
4.86
6.18
4.87
7.34
4.71
6.80
Corpus Christi
4.66
5.98
4.78
6.48
4.57
5.88
Downing
4.61
6.94
4.57
6.47
4.38
6.65
Emmanuel
4.56
5.81
4.75
6.28
4.58
5.93
Fitzwilliam
5.64
6.31
5.28
6.25
5.47
6.21
Girton
4.74
5.79
4.38
6.24
4.60
5.78
Gonville & Caius
4.57
6.68
4.65
6.67
4.54
6.43
Homerton
5.01
6.42
4.69
6.59
4.87
6.33
Hughes Hall
6.15
—
5.65
—
5.52
—
Jesus
4.30
6.42
4.39
6.65
4.28
6.38
King's
4.34
5.82
4.34
5.63
4.48
5.80
Lucy Cavendish
5.18
6.14
5.01
6.10
5.07
6.03
Magdalene
5.26
6.05
5.27
6.35
5.30
6.22
Murray Edwards
5.06
7.08
4.50
6.64
4.94
6.95
Newnham
4.61
6.40
4.54
5.95
4.67
5.88
Pembroke
4.59
6.38
4.61
6.46
4.42
6.48
Peterhouse
5.00
6.41
4.84
5.92
4.97
6.11
Queens'
4.82
6.58
4.64
6.57
4.78
6.40
Robinson
5.72
6.54
5.37
6.58
5.58
6.99
Selwyn
4.58
6.23
4.51
7.06
4.47
5.79
Sidney Sussex
5.28
5.98
5.09
6.02
5.19
5.97
St Catharine's
4.75
5.60
4.74
5.86
4.51
5.70
St Edmund's
6.84
5.93
6.37
5.79
6.49
5.98
St John's
4.38
5.84
4.41
6.52
4.25
5.85
Trinity
4.61
6.36
4.40
5.93
4.66
6.24
Trinity Hall
5.15
6.75
4.86
7.00
5.15
6.64
Wolfson
4.33
6.45
3.92
6.25
5.28
6.53
⚠ Interpreting St Edmund's: St Edmund's shows applicant averages above offer holder averages across all three sections. This likely reflects a very small, mature-entry applicant pool where pool-in offers from other colleges are attributed to St Edmund's in the FOI data, creating a counterintuitive pattern. Wolfson similarly has a compressed score range due to its graduate-entry focus. Treat these two colleges as outliers when interpreting the chart.
Key findings · Score targets · What the data cannot tell us
S1 (Maths 1) is the dominant ESAT predictor. Offer rates climb steeply with S1 for Home applicants — from 2% at band 1–2 to 100% at band 9.
Finding 2
Chinese applicants need systematically higher ESAT scores for the same offer probability. The average S1 of Chinese offer holders (7.95) is 2.72 points above that of Home offer holders (5.23).
Finding 3
The pool conversion gap reveals the interview stage as the source of the domicile difference. Pool conversion is 32.8% for UK applicants and only 5.3% for Chinese applicants — a 6× gap driven by interview-stage factors.
Finding 4
A-level grades are a floor, not a differentiator. A*A*A*A* is the most common predicted grade string for both offer and non-offer holders. The ESAT and interview are doing the actual selection work.
Finding 5
Cambridge offers to Chinese applicants happen at very high overall scores. Chinese offer holders average a combined ESAT total of 23.8 out of 27, while Home offer holders average 16.0.
Practical targets for applicants
Home Engineering: S1 ≥ 6.0 for a majority offer rate; S1 ≥ 7.0 puts you in the near-certain range
Home Engineering total: aim for combined ≥ 16–17 out of 27 (offer-holder mean)
China Engineering: no S1 band guarantees >50% offer rate; aim for S1 = 9.0 and total ≥ 22–24
China Engineering: at S1 ≥ 8, offer rate is 35.7% — still a minority, requiring strong interview performance
International Engineering: aim for S1 ≥ 7.0; total ≥ 19–21 puts you near International offer-holder mean
Natural Sciences (all domiciles): Physics S1 is the strongest predictor for Physical NatSci; aim for ≥ 7.0 in Physics
Grades: A*A*A is the published minimum — but so do most rejected applicants
Interview: the pool conversion data shows interviews are decisive — ESAT alone does not secure an offer at any score band
Critical limitation — no interview scores: Cambridge refused to provide interview data in all FOI responses. The interview is the final and decisive stage, and on the basis of these records we cannot determine what interview scores look like, whether they correlate with ESAT scores, or how they are weighted.
All information current as of May 2026. Cambridge admissions criteria, ESAT requirements, and domicile quota structures can change between cycles. Always verify on the Cambridge Admissions website and check the current UAT-UK course list before applying.
Enjoy Reading This Article?
Here are some more articles you might like to read next: