Fast Stream Acceptance Rate 2026: 69,046 Applications, 703 Recommended

The Fast Stream acceptance rate, as the government measures it, was 1.0% for the 2026 intake: 703 people were recommended for appointment from 69,046 applications. That is about one person recommended for every 98 who applied (our arithmetic). Two words matter in that sentence. GOV.UK does not call it an acceptance rate: it publishes a success rate, defined as people recommended for appointment divided by applications by first preference. And no official page publishes an offers-accepted rate, so how many of those recommended took up a place is not public. So how hard is the Fast Stream? Very, on any published measure. But "about 1%" hides as much as it tells you. The rate for a first-preference scheme ran from 0.2% to 4.7%, and it moves with your age, your university, whether you already work in the Civil Service and your family background. This guide shows those official rates one by one, with a tool that puts the figures for someone like you next to the overall rate, and is plain about the one thing the data cannot do: give you a personal probability.

Civil Service Fast Stream recruitment data 2026, Table 1; Civil Service Careers, Understanding the application process (read 16 October 2026).

Applications for 2027 entry are open from midday 8 October 2026 to midday 5 November 2026, according to the official application process page. If you are deciding whether it is worth applying, the short version of this page is: the odds are long for everyone, they are not the same for everyone, and the first gate, the online tests, is one every applicant faces.

The competition roughly doubled in difficulty in one year and then stayed there. The 2024 intake had a success rate of 2.2%; for 2025 and 2026 it was 1.0% and 1.0%. Two things moved at once: first-preference applications rose from 44,362 to 72,691, and vacancies fell from 915 to 676. The longer series back to 2013, and each scheme's rate year by year, are in our Fast Stream success rates by scheme guide, so they are not repeated here. This page goes the other way: from the whole field down to the groups inside it.

So the published numbers show only the start and the end: applications in, people recommended out. How many fall at the online tests, how many at the assessment centre, and how many at a Final Selection Board is not in any official page we could find. Be wary of any figure that claims to be a Fast Stream "online test pass rate": it is not from the government's data. What the official pages do make clear is the shape of the funnel. The online tests are the one stage every applicant must pass, and the online-tests page itself warns that your feedback "does not indicate whether you will progress to the next stage, as the process is highly competitive".

That is the stage you can prepare for most directly. The Work-based scenarios test asks you to rate how effective actions are on a scale that runs Counterproductive, Ineffective, Partially effective, Effective; the Case Study uses the same scale in some situations and asks for the most and least effective response in others. Our Fast Stream online tests guide explains the format with the official worked example, and the ProSyllabus Fast Stream practice set gives you original scenarios in the same rating format, with the reasoning for every answer, to practise the first gate before you sit it.

It is tempting to take your scheme's rate, your university's rate and your age group's rate and blend them into a single "my chance" number. The data does not allow it, for a simple reason: every table is one-dimensional. Table 5 gives success by university; Table 7 gives success by age. Nothing tells you the success rate of, say, 21 to 23 year olds from Oxford or Cambridge applying to Government Policy. Those groups overlap heavily, so their rates are not independent, and multiplying or averaging them would count the same people several times over.

There is a second reason. A scheme's rate is measured on applicants who put that scheme first, while the age, university and background tables pool applicants from every scheme. Your scheme figure and your age figure are describing two different crowds. The honest reading is the one the tool gives: each figure tells you how a group did, and you can see which way each part of your profile leans, but not how far they move you together.

Your first-preference scheme is the one characteristic on this page you choose. In 2026 the highest rate was the Government Economics Service, at 4.7%, and the lowest was Diplomatic and Development (incl. D&D Economics), at 0.2%: 13,550 first-preference applications for 25 vacancies. The scheme rates for each year are in our success rates by scheme guide. The table below adds two things that guide does not: the number of applications for every person recommended, which is easier to picture than a percentage, and the split between external and internal applicants.

Read the per-person column as a ratio of crowd to winners. In 2026, Government Policy, the largest scheme with 201 vacancies, recommended one person for every 48 applications. Diplomatic and Development recommended one for every 645. Economics recommended one for every 21, Operational Research one for every 31 and Statistics one for every 38. Each scheme sets its own eligibility, so check the scheme pages (our schemes and eligibility guide collects them) before reading a low ratio as an open door. A small scheme can swing a lot from year to year: Risk Management had 5 vacancies in 2026 and 11 in 2025, so a handful of people changes its rate.

It surprises many people, but applicants already inside the Civil Service did worse than outsiders in each of the last three intakes. In 2026 the internal success rate was 0.6% (66 recommended from 11,125 applications), against 1.1% for external applicants. In 2025 the figures were 0.6% and 1.1%; in 2024, 1.7% and 2.3%.

The data does not explain the gap, and we will not invent a reason. Two facts from the official in-service route page are worth knowing, though. Serving civil servants can apply to 12 Fast Stream options without a degree (the page lists Commercial, Cyber Security, Digital, Diplomatic & Development, Fast Stream Yorkshire and the Humber, Finance, Government Policy, HR, Operational Delivery, Project Delivery, Property and Risk Management), so the internal field is not drawn from the same population as the external one. And staff at SEO grade or above who succeed "must accept a salary reduction" into the Fast Stream grade. Our reading: the internal pool is wider in background and the data cannot say which part of it does well.

Inside the internal pool, age matters in the same direction as outside it. Internal applicants aged 21 to 23 were recommended at 1.4% in 2026, and those aged 30 to 39 at 0.4% (Table 12).

The age table is the clearest gradient in the data. In 2026, applicants aged 21 to 23 were recommended at 1.4%, and the rate fell in every older band, to 0.2% for those aged 40 and over: 11 people from 5,487 applications. The 21 to 23 band was also by far the largest, with 28,010 applications. From the 21 to 23 band upwards, the rate fell in every older band in 2025 and 2024 as well. The youngest band, 20 or under, is the exception that moves: it was the highest of all in 2025, at 1.8%.

Where you took your first degree shows the widest gap of any background question. In 2026, applicants whose undergraduate university was Oxford or Cambridge were recommended at 4.4%, applicants from the rest of the Russell Group at 1.6%, and applicants from any other provider at 0.5%. On those two published rates, the Oxbridge figure is about 8.8 times the "Other education provider" figure (our arithmetic). But look at the counts too: "Other education provider" applicants were the largest group, with 39,497 applications, and 197 of them were recommended, which is more people than the 101 from Oxford and Cambridge.

Degree subject made little difference. STEM graduates were recommended at 1.0% and non-STEM graduates at 1.1% in 2026. GOV.UK defines the Russell Group as 24 UK universities including Oxford and Cambridge, and STEM by the Higher Education Statistics Agency's science marker or qualification types such as BSc, MSc and Eng.

Home region matters less than university or age. In 2026 the highest published regional rates were London and the South East, both 1.3%; the lowest was Wales, at 0.6%. Northern Ireland, the Channel Islands and the Isle of Man each had fewer than five people recommended, so GOV.UK does not publish their rates. London alone supplied 16,696 applications. There is no external or internal split by region.

The socio-economic tables point the same way on every measure. In 2026, applicants who attended an independent or fee-paying school without a bursary were recommended at 2.0%, against 1.0% for non-selective state schools. Applicants eligible for free school meals did worse (0.7%) than those who were not (1.3%). So did applicants whose parents had no formal qualification (0.6%) against those with a degree-level parent (1.3%), and applicants who described themselves as from a lower socio-economic background (0.7% against 1.3%).

Two cautions. First, many applicants did not give a usable answer: 16,665 applications were in the “Don’t know/prefer not to say/non-respondent” row for school type, and those applicants were recommended at 0.6%. Second, the measures overlap with each other and with university, so the gaps are not five separate effects. For internal applicants GOV.UK publishes its own table (Table 13), and in 2026 the gaps there were smaller; the tool uses it when you say you are a civil servant.

Whatever stage you are at, getting this far takes real effort, and that is worth being proud of. Prepare at a pace that works for you, check the official Civil Service Careers pages for anything that decides your application, and look after yourself along the way. We are wishing you every success.