What Is the ONS Secure Research Service — and Can a Student Use It? (2026)

The Office for National Statistics holds de-identified, unpublished microdata — individual-level records from surveys, the Census and administrative systems — that never appear on any download portal. Access runs through the Secure Research Service (SRS), and it is gated by a formal accreditation regime rather than a registration form. This guide answers the questions students actually ask: what the SRS is, who counts as an accredited researcher, whether a master’s or PhD student can realistically get in, how long it takes, and what to use instead when the answer is no.

What is the Secure Research Service?

The SRS is the ONS’s trusted research environment: it “gives accredited researchers secure access to de-identified, unpublished data in order to work on research projects for the public good” (ONS). It operates under the Digital Economy Act 2017 and the Statistics and Registration Service Act 2007, and it is built around the Five Safes framework — a set of principles adopted by secure labs covering safe people, safe projects, safe settings, safe data and safe outputs. In practice that means: you must be accredited (people), your project must be separately accredited as serving the public good (projects), you work inside a controlled environment rather than downloading files (settings), the data are de-identified (data), and anything you take out is checked before release (outputs).

This is a different world from ordinary secondary data analysis. Most students working with UK survey data never need it — the safeguarded tier of the open data repositories ecosystem covers the standard research datasets. The SRS exists for the detailed, sensitive versions: finer geography, more granular occupational codes, linked administrative records.

Who counts as an accredited researcher?

Accreditation is personal, not institutional, and it has two routes (ONS, “Become an accredited researcher”):

The provisional route carries three conditions that matter for students: the supervising accredited researcher must work for the same organisation as you, must agree to oversee all your projects, and must accept responsibility if you breach your access declaration. Acceptable degree content includes ethics, data analysis, research design and statistical testing; relevant dataset experience includes administrative, census, cohort, longitudinal and survey data.

So can a student actually use it?

Follow the criteria to their logical conclusion:

• PhD and research master’s students: yes, realistically. A quantitative degree at 2:2 or above satisfies the full-accreditation education test, and if it does not, the provisional route works where your supervisor (or another researcher at your university) holds full accreditation and accepts the supervision conditions.
• Taught master’s students: possible, but the timeline is the constraint — see below. The project itself must also be accredited as research for the public good, which a three-month dissertation can struggle to justify.
• Undergraduates: rarely. Full accreditation requires a completed degree or three years of quantitative research experience, so the provisional route (with a fully accredited supervisor at your institution) is effectively the only door, and the project-accreditation and training requirements usually make a final-year project impractical.

The scale of the scheme is public: as of July 2026 there were 6,406 current accredited researchers, of whom 1,596 were actively involved in live projects, with 1,171 new researchers starting the accreditation process in the previous 12 months (ONS management information).

How long does accreditation take?

The ONS publishes its own processing figures: researcher accreditation takes a mean of 32 working days (median 24) from submitting your documents to accreditation being granted. That is only one stage — you also need a Hub account (the ONS aims to respond to account applications within five working days, and you must activate two-step verification within 24 hours of the welcome email), completion of the required training, and separate accreditation of your research project. Budget a term, not a fortnight, between deciding you need secure data and running your first analysis. If your dissertation deadline cannot absorb that, design around the safeguarded tier instead — the difference is explained in our guide to registering with the UK Data Service.

Where do you physically work with the data?

You do not download SRS data; you connect to it. Routes into secure environments in the UK include institutional access arrangements and physical safe settings. The UK Data Service’s access-conditions page documents the equivalents for its own SecureLab: connection from an approved institutional device, SafePods in the SafePod Network, or the UK Data Archive Safe Room at the University of Essex — which can also host access to the ONS SRS itself for users holding an Accredited Researcher Assurance Registration signed by both organisations. The practical point for planning: secure access is tied to approved places and devices, so check what your university offers before committing your methodology chapter to secure data.

Researcher accreditation is only half the gate

A common planning error is to treat accreditation as one hurdle when it is two. Your researcher accreditation certifies you; your project must be separately accredited before you touch data. The project application is where public-good justification lives, and it is also where a vague dissertation plan gets exposed: you will need to state what data you need, why the research serves the public good, and what outputs you intend to produce. Outputs themselves are checked before they leave the environment — a table of regression coefficients will pass; anything that could disclose an individual will not, and the checking adds days to your write-up timeline that most dissertation Gantt charts never include.

Plan the sequence in this order: confirm a fully accredited supervisor exists at your institution (for the provisional route), draft the project application with your supervisor’s input, submit researcher and project applications as early as your programme allows, and complete the mandatory training while they process. Students who start this in the semester before their dissertation term are the ones who actually analyse secure data; students who start in week one of the dissertation usually end up on the safeguarded tier anyway, minus the lost weeks.

What to use instead if you don’t need the secure tier

Most dissertation questions are answerable one tier down. Safeguarded versions of the major UK surveys — downloadable after registration under an End User Licence — carry the same variables at coarser detail, and open data need no registration at all. The decision is a design question: if your analysis needs local-authority-level geography or detailed occupation codes, the secure tier is the honest answer and your timeline must absorb it; if regional breakdowns will do, the safeguarded file gets you writing months earlier. Whether that trade-off suits a desk-based dissertation design is worth settling before you commit.

Frequently asked questions

What data can you access through the SRS?

De-identified, unpublished microdata held by the ONS for accredited research — the detailed versions of surveys, Census data and linked administrative data that are too disclosive to publish or distribute under ordinary licences. The ONS maintains a “Discover secure research data” catalogue of what it holds.

Is accreditation free?

The ONS accreditation process itself is applied for through the Research Accreditation Service Hub; the real costs are time (mean 32 working days for researcher accreditation alone) and mandatory training, not fees.

What are the Five Safes?

Safe people, safe projects, safe settings, safe data, safe outputs — the framework the SRS uses to make secure data use auditable at every layer, from who you are to what leaves the environment.

Does my dissertation count as a “public good” project?

Project accreditation is a separate application from researcher accreditation, and your project must demonstrate its public benefit. Academic research can qualify, but “it’s for my degree” is not itself the case — write the application around the research contribution.

What is the Integrated Data Service?

The IDS is the ONS’s cloud-based secure infrastructure supporting analysis for statistical priorities — a newer platform in the same governed family. For a student, the entry point remains researcher and project accreditation.

Do I need ethics approval as well?

Accreditation does not replace your university’s ethics process — secure-data projects still go through institutional review. See our guide to whether you need ethical approval.

Design the chapter before the application

The worst time to discover a 32-working-day accreditation queue is week two of your analysis phase. Map your data tier, your timeline and your fallback dataset before you write the methodology. Tesify helps you plan those chapters and keep the design decisions documented as you write — 100% written by you.

Plan your dissertation with Tesify