Key finding: Only 33% of South Africa’s 2018 first-time entering undergraduate cohort graduated within the regulation time of their qualification, and 62% had graduated within two years beyond it, according to the Council on Higher Education’s VitalStats 2023 (published 2025, excluding Unisa). At postgraduate level the picture inverts: 83% of the 2018 Honours cohort graduated by 2023, but only 64% of the coursework master’s cohort did — with 36% having dropped out.
Where these numbers come from — and what they are not
The figures below are cohort throughput statistics from VitalStats: Public and Private Higher Education 2023, published by the Council on Higher Education (CHE) in 2025, drawing primarily on the Department of Higher Education and Training’s HEMIS audited datasets. A cohort study follows the students who entered a qualification in a given year — here, 2018 — and counts who graduated or dropped out in each subsequent year through 2023.
Two warnings before you quote anything. First, these are not the same measure as the “graduation rate” you get by dividing one year’s graduates by the same year’s enrolments — that ratio mixes different cohorts and is routinely mislabelled as a completion rate; our postgraduate enrolment statistics guide covers that trap in detail. Second, the cohort figures below exclude Unisa unless stated: the CHE analyses Unisa separately because large numbers of its students legitimately interrupt their studies, which distorts pooled cohort mathematics. On top of that, the CHE marks 2023 figures with the caveat that potential graduates may remain in the system after 2023 — so the final-year numbers are floors, not ceilings.
Undergraduate throughput: the 2018 cohort
| Qualification (2018 entrants, excl. Unisa) | Graduated in regulation time | By regulation +1 year | By regulation +2 years | Source |
|---|---|---|---|---|
| 360-credit diplomas | 22% | 45% | 57% | CHE VitalStats 2023, Fig. 180 |
| 3-year degrees | 30% | 52% | 59% | CHE VitalStats 2023, Fig. 180 |
| 4-year degrees | 46% | 64% | 69% | CHE VitalStats 2023, Fig. 180 |
| Weighted national rate | 33% | 54% | 62% | CHE VitalStats 2023, Fig. 180 |
Each statistic in one sentence, quotable on its own:
- 22% of the 2018 diploma cohort graduated in regulation time; 57% had graduated within two extra years (CHE VitalStats 2023).
- 30% of the 2018 three-year degree cohort graduated in regulation time; 59% within two extra years (CHE VitalStats 2023).
- 46% of the 2018 four-year degree cohort graduated in regulation time; 69% within two extra years (CHE VitalStats 2023).
- The comparison across the 2016, 2017 and 2018 cohorts is remarkably stable: overall regulation-time graduation of 34%, 33% and 33% respectively, and 59%, 62% and 62% by two years beyond (CHE VitalStats 2023, Fig. 181).
Dropout: the other side of the same cohort
By 2023, 38% of the 2018 diploma cohort had dropped out (26% by 2020, rising through 31% and 35%), against 62% graduated (CHE VitalStats 2023, Fig. 145). For 4-year degrees the accumulated dropout reached 31% by 2023, against 69% graduated (Fig. 153). Remember the CHE’s methodological note: students who interrupt and later return (“stop-outs”) are credited back in the year they return, so accumulated dropout can even fall slightly between years — dropout here means left the qualification without completing it during the window, not “gone forever”.
The gender gap is one of the sharpest patterns in the cohort data. Among 2018 entrants to 3-year degrees, 71% of women had graduated by 2023 against 52% of men — and the accumulated dropout was 29% for women against 48% for men (CHE VitalStats 2023, Fig. 149). Nearly half the men who started a 3-year degree in 2018 had left without it five years later.
Postgraduate throughput: Honours holds, master’s leaks
| Qualification (2018 entrants, excl. Unisa) | Graduated, first year(s) | Graduated by 2023 | Dropped out by 2023 | Source |
|---|---|---|---|---|
| Postgraduate diplomas/certificates | 60% in first year | 84% | 16% | CHE VitalStats 2023, Fig. 188 |
| Honours | 63% in first year | 83% | 17% | CHE VitalStats 2023, Fig. 193 |
| Coursework master’s | 44% within ~2 years (2020) | 64% | 36% | CHE VitalStats 2023, Fig. 194 |
The quotable versions:
- 83% of the 2018 Honours cohort (excluding Unisa) had graduated by 2023; 63% finished in the first year (CHE VitalStats 2023).
- Only 64% of the 2018 coursework master’s cohort had graduated by 2023 — the accumulated total moved 44% → 54% → 61% → 64% across 2020–2023 — while accumulated dropout reached 36% (CHE VitalStats 2023).
- Institutional variation is wide: the share of the 2018 Honours cohort graduating within three years ranged from 63% to 95% across universities, and for postgraduate diplomas from 48% to 99% (CHE VitalStats 2023, Figs. 189, 191).
That master’s number deserves a moment. More than one in three students who started a coursework master’s in 2018 had dropped out by 2023 — a substantially higher attrition than Honours (17%) or postgraduate diplomas (16%). The structural difference is the dissertation component: coursework is schedulable, a mini-dissertation is not, and the data shows exactly where momentum dies. It is also why the master’s completion curve keeps climbing years after entry: 10% of the cohort graduated in 2021 alone, 6% in 2022, 3% in 2023 — finishing late is normal, and the students who stay enrolled keep converting.
How to read a cohort table without fooling yourself
Three reading rules keep these numbers honest in your own writing. First, always name the cohort: “the 2018 entering cohort” is a specific group of people, and their 62% is not a property of South African universities in general — it is a property of that cohort’s journey through 2023. Second, distinguish accumulative from non-accumulative figures: the accumulative number tells you where the cohort stands in total by a given year; the non-accumulative number tells you what happened in that year alone. The master’s series above shows why it matters — the 64% headline hides that 10% of the whole cohort graduated in 2021 alone, which is the single most encouraging fact in the table for anyone currently behind schedule. Third, respect the asterisk: the CHE marks the final year because students still enrolled can still graduate, so every “by 2023” figure understates the cohort’s eventual completion. A dissertation that quotes these numbers with their caveats reads like the work of someone who handled the source; one that quotes “SA dropout rate is 38%” reads like someone who handled a headline.
The same discipline applies to the institutional ranges. The 63%–95% spread on Honours completion is not a league table you can rank universities with — institutions differ in intake profile, programme mix, and the share of part-time students, and the cohort method treats a student who switched qualifications as a dropout from the first one. The spread is evidence that institutional context matters enormously; it is not evidence that the 95% institution teaches better than the 63% one.
What this means if you are the statistic
The cohort data carries two practical messages. First, regulation time is the exception, not the norm — at every undergraduate level, most students take longer, and the system’s own published figures say so. Budget your registration years accordingly rather than treating an extra year as personal failure. Second, at master’s level the risk is not failing the dissertation; it is drifting out of it. The one-in-three attrition happens quietly, semester by semester. Structure is the countermeasure: a chapter plan with dates, visible progress, and a writing system that survives your job — which is what Tesify exists to provide, and what our guides to writing while working full-time and choosing between Unisa and contact universities are built around.
FAQ
What is South Africa’s university dropout rate?
There is no single national dropout rate. The most defensible figures are cohort-specific: for 2018 entrants (excluding Unisa), accumulated dropout by 2023 was 38% for 360-credit diplomas, 31% for 4-year degrees, 17% for Honours and 36% for coursework master’s (CHE VitalStats 2023).
What is a throughput rate?
The share of a first-time entering cohort that has graduated by a given point — in regulation time, or regulation plus one or two years. It is calculated against the cohort’s baseline enrolment, not against a later year’s headcount.
Why is Unisa excluded?
The CHE analyses Unisa separately because many of its students interrupt their studies for years, which breaks pooled cohort assumptions. Unisa-specific cohort figures are published separately in VitalStats.
Is the 33% regulation-time figure getting worse?
No — it is stable. The 2016, 2017 and 2018 cohorts show 34%, 33% and 33% respectively, and 59–62% by two years beyond regulation time (CHE VitalStats 2023).
Do more women than men finish their degrees?
In the 2018 three-year degree cohort, decisively yes: 71% of women had graduated by 2023 versus 52% of men, with male accumulated dropout at 48% versus 29% for women (CHE VitalStats 2023).
How many master’s students drop out in South Africa?
Of the 2018 coursework master’s cohort (excluding Unisa), 36% had dropped out by 2023, while 64% had graduated (CHE VitalStats 2023). Both figures may still shift slightly as remaining students complete.
Where can I get the underlying data?
VitalStats 2023 is a free CHE publication; the underlying enrolment data sits in DHET’s HEMIS. For survey microdata routes, see our guide to accessing South African data for your dissertation.
Can I cite these figures in my dissertation?
Yes — cite the CHE publication with the figure number and the cohort year, state that figures exclude Unisa, and keep the 2023 caveat that late completers may still be in the system.