Key finding: 85,6% of South African households had at least one member with access to the internet by any means in 2025 — but only 20,6% had fixed internet at home, according to Statistics South Africa’s General Household Survey 2025 (statistical release P0318, published 26 May 2026). For students, the gap between those two numbers is the story: connectivity in South Africa overwhelmingly means a mobile connection, paid for by the gigabyte.
The national picture: access exists, infrastructure lags
The General Household Survey is Stats SA’s annual household survey; the 2025 release is the most recent, and its internet tables (Table 14.1) are unusually quotable. Each figure below is stated so it can be cited on its own, with source and year.
| Type of internet access, households, 2025 | National (RSA) | Highest province | Lowest province | Source |
|---|---|---|---|---|
| Any kind of access | 85,6% | Western Cape 93,8% | Eastern Cape 74,5% | Stats SA GHS 2025, Table 14.1 |
| Mobile access | 78,9% | KwaZulu-Natal 84,9% | Northern Cape 68,8% | Stats SA GHS 2025, Table 14.1 |
| Fixed internet at home | 20,6% | Western Cape 49,7% | Limpopo 5,9% | Stats SA GHS 2025, Table 14.1 |
| Internet at work | 12,9% | Western Cape 21,5% | Limpopo 3,7% | Stats SA GHS 2025, Table 14.1 |
| Public Wi-Fi | 8,2% | Western Cape 16,4% | Limpopo 2,3% | Stats SA GHS 2025, Table 14.1 |
| At an educational facility | 4,0% | Western Cape 7,6% | Limpopo 0,7% | Stats SA GHS 2025, Table 14.1 |
| At a library | 2,1% | Western Cape 6,8% | Northern Cape 0,2% | Stats SA GHS 2025, Table 14.1 |
The quotable sentences:
- 85,6% of South African households had at least one member who could access the internet at one or more locations in 2025 (Stats SA, GHS 2025).
- Only 20,6% of South African households had fixed internet access at home in 2025 — from 49,7% in the Western Cape down to 5,9% in Limpopo (Stats SA, GHS 2025).
- 78,9% of households could access the internet using mobile devices in 2025 (Stats SA, GHS 2025).
- Households in metropolitan areas had 89,3% internet access, against the 85,6% national figure; fixed home access in the metros was 33,5% against 20,6% nationally (Stats SA, GHS 2025, Table 14.2).
- Access through an educational facility (4,0%) and libraries (2,1%) remains marginal nationally (Stats SA, GHS 2025).
What the provincial spread means for a distance student
The provinces where fixed home internet is rarest — Limpopo at 5,9%, Mpumalanga at 7,3% — are exactly where distance study is often the only realistic route to a postgraduate degree. A student in Polokwane weighing Unisa against a contact university is, statistically, planning to write a dissertation over a mobile connection: the practical difference between the 85,6% headline and the 20,6% fixed-line reality is that “having access” means paying per gigabyte, managing signal, and treating every large download as a decision. Fieldwork platforms, video supervision meetings, cloud backups and literature databases all consume data — and the GHS numbers explain why data cost discipline features so prominently in every realistic account of writing a dissertation around a full-time job in South Africa.
The education context in the same survey
The GHS 2025 also measures who is in the education system at all: 30,7% of individuals aged five and older attended an educational institution in 2025; of those attending, 87,7% were in primary or secondary schools, 5,4% at tertiary institutions, and 2,5% at TVET colleges (Stats SA, GHS 2025, Table 4.2). Read alongside the connectivity tables, the policy question writes itself: the small tertiary cohort is drawn disproportionately from households on the wrong side of the fixed-line divide, and the national infrastructure through which they could study data-free — libraries at 2,1%, educational facilities at 4,0% — barely registers as an access channel.
For sector-level education statistics — enrolment, graduation and throughput — the GHS is the wrong instrument: those come from HEMIS via the DHET and CHE publications, and we cover them in our postgraduate enrolment statistics and dropout and throughput statistics guides. The GHS answers a different question: what the households students live in actually have.
Using these figures inside your own research design
Beyond citing them as context, the connectivity tables should actively shape method decisions in South African studies, and examiners notice when they do. If your design assumes participants will complete an online survey, the 20,6% fixed-access figure is a sampling-bias warning in official form: an online-only instrument systematically under-reaches the households your Limpopo or Eastern Cape sample lives in, and your limitations section should say what you did about it — mobile-first questionnaire design, data-free platforms, or a paper alternative. If your study is about digital anything — e-learning, e-government, fintech adoption — the gap between any-access (85,6%) and fixed access (20,6%) is not background; it is a variable, and the provincial spread gives you a defensible basis for site selection. And if you are interviewing participants over video, the tables predict your no-show pattern better than any pilot: a participant whose only connection is prepaid mobile data is rationing it, and offering a voice-call or in-person alternative is both methodologically sounder and more ethical than treating dropped connections as non-compliance.
The same numbers also justify research-administration choices to your supervisor. Scheduling supervision calls as voice rather than video, batching document exchanges, or asking the department to print consent forms are not signs of a disorganised student — they are evidence-based adaptations to a documented infrastructure reality, and the GHS citation turns each one from an excuse into a design decision.
How to cite and how not to misread these figures
Three cautions, because these numbers are widely misquoted. First, the unit is the household, not the person: “85,6% of households had a member with access” is not “85,6% of South Africans have internet”. Second, the access types are not exclusive — a household can appear in the mobile, work and public Wi-Fi rows simultaneously, so the rows must never be summed. Third, the measure is access or use by at least one member at one or more locations — a generous definition that says nothing about speed, reliability or affordability of use for a second household member. A dissertation citing these figures states the definition; a content farm does not. The reference, in Harvard form:
Statistics South Africa. 2026. General Household Survey 2025. Statistical release P0318. Pretoria: Statistics South Africa.
If you want to go deeper than the published tables — say, internet access cross-tabulated by province and age for your own methodology chapter — the microdata route through SuperWEB2 and DataFirst is free and documented in our step-by-step data access guide.
Why the household unit matters more than it looks
One definitional subtlety deserves its own paragraph, because it changes what the headline means for students specifically. The GHS counts a household as connected if at least one member had access at one or more locations — home, work, place of study, an internet café or a public hotspot. In a six-person household where one member has internet at work, the household is “connected”, while the postgraduate student in that same household may have no practical access for an evening literature search. The 85,6% figure is therefore an upper bound on student-usable connectivity, not an estimate of it; the figures closer to a student’s lived reality are fixed home access (20,6%), mobile access (78,9%, constrained by data cost), and the near-absence of library access (2,1%). Writing this distinction into your dissertation — or your funding motivation — is exactly the kind of careful measurement reading that separates a citable claim from a headline, and it is why serious South African studies quote the definition alongside the number.
What this means for how you set up your dissertation workflow
If your household is in the 79% mobile-first majority, design the workflow around scarcity: draft locally and sync deliberately, batch your literature downloads on campus or workplace Wi-Fi, and keep your working documents in a single project rather than scattered across attachments that each cost data to re-fetch. This is a real reason cloud-light tools matter in the South African market: Tesify keeps your chapters, sources and reference list in one place, so progress made in a lunch-break session on campus Wi-Fi is still there on the taxi home — without a gigabyte of sync traffic behind it.
FAQ
What percentage of South Africa has internet access?
85,6% of households had at least one member with access to the internet by any means in 2025 (Stats SA, GHS 2025). Note the unit: households with any access, not individuals with personal access.
How many South African households have Wi-Fi or fibre at home?
20,6% of households had fixed internet at home in 2025 — dial-up, ADSL or fibre — ranging from 49,7% in the Western Cape to 5,9% in Limpopo (Stats SA, GHS 2025).
Is mobile data the main way South Africans get online?
Yes. 78,9% of households could access the internet via mobile devices in 2025, against 20,6% with fixed home access (Stats SA, GHS 2025).
Which province has the best internet access?
The Western Cape, at 93,8% of households with any access and 49,7% with fixed home internet in 2025; the Eastern Cape (74,5% any access) and Limpopo (5,9% fixed) anchor the other end (Stats SA, GHS 2025).
How many students study over mobile connections?
The GHS measures households, not student devices, so no exact figure exists in this release. The defensible inference: with 20,6% fixed home access and 78,9% mobile access, most students outside the metros connect primarily through mobile data.
Where does this data come from?
Statistics South Africa’s General Household Survey 2025, statistical release P0318, published 26 May 2026 — Tables 14.1, 14.2 and 4.2. It is a free download from Stats SA.
Can I use these figures in my dissertation?
Yes — cite the release with its series number and reference year, keep the household unit explicit, and do not sum the access-type rows. For deeper cross-tabulations, use the microdata route.
Is internet access improving?
The GHS 2025 release describes access as having continued to expand rapidly, reaching 85,6% of households by 2025, with fixed access increasing steadily to 20,6%. Cite trend claims to the release’s own narrative rather than reconstructing them across editions with different methodologies.