Roughly 25 to 30 percent of UK Global Talent Visa endorsement applications are approved at first attempt. That figure comes from FOI-derived data (OTB Legal), which is the only public source for endorsement-stage rejection rates. Once an applicant clears endorsement, the visa stage itself approves in well over 90 percent of cases. The bottleneck is endorsement, not the visa.
If you have read elsewhere that the success rate is "85 to 93 percent," that number is real but it refers only to the visa stage — after endorsement is already in hand. The endorsement stage itself, where most applications fail, has a far lower public approval rate.
This article unpacks both numbers, where they come from, and what they mean for your application strategy.
Endorsement vs visa stage: two different success rates
The UK Global Talent Visa runs on a two-stage model.
Stage 1 — Endorsement. You submit evidence to one of the six endorsing bodies (Tech Nation, UKRI, Royal Society, British Academy, Royal Academy of Engineering, Arts Council England). They decide whether your record meets the bar for "exceptional talent" or "exceptional promise." This is the hard gate.
Stage 2 — Visa application. With the endorsement in hand, you apply to the Home Office for the actual visa. The Home Office checks identity, criminal history, biometrics and immigration history. Almost everyone who clears Stage 1 clears Stage 2.
When you see "Global Talent Visa success rate 90 percent," it almost always refers to Stage 2 outcomes only. The number is technically correct but it ignores the funnel above it.
What the FOI data actually shows
Most public discussion of endorsement rejection rates traces back to a Freedom of Information request (OTB Legal) that pulled endorsement outcomes from the Home Office and the endorsing bodies. The headline finding: the cumulative endorsement rejection rate sits in the 70 to 75 percent range across recent years.
That number is high. It is also broadly consistent with what experienced practitioners report and with the volume of forum threads describing failed first attempts.
A few caveats:
- The data is not split cleanly by endorser, discipline or year. Public per-endorser rates are not officially published.
- Some endorsing bodies handle very low absolute volumes (under 100 applications a year), so single-year per-body percentages move dramatically.
- The 70 to 75 percent rejection rate is endorsement only. It does not capture withdrawn applications, applications that never submitted after a free Endorsement Review told the applicant they weren't ready, or repeat submissions.
What the data is good for: a realistic baseline expectation. What it is not: a precise cohort split.
Per-endorser approximations
Endorsing bodies do not publish their own approval rates. Practitioners working across multiple bodies report a rough hierarchy that mirrors the FOI data and the volume distribution.
| Endorser | Discipline | Approximate first-attempt approval | Volume signal |
|---|---|---|---|
| Tech Nation | Digital technology | ~25–30% | High volume, sharper bar since Jan 2025 |
| UKRI | Research / academic | ~40–50% | Lower volume, stricter on grant evidence |
| Royal Society | Natural / medical sciences | ~30–40% | Peer-review heavy |
| British Academy | Humanities / social science | ~30–40% | Peer-review heavy |
| Royal Academy of Engineering | Engineering | ~30–40% | Lower volume |
| Arts Council England | Arts, film, fashion, architecture | ~25–35% | Mixed evidence formats |
These are practitioner approximations from cumulative observation, not official figures. We publish them with that caveat because no endorser publishes its own data and applicants need a starting frame.
The pattern: digital technology has the largest applicant pool and the toughest pass rate. Research routes (UKRI, learned societies) are smaller and slightly more forgiving when the evidence aligns with academic norms.
Talent vs Promise: success rate is not where the difference shows up
A common question: do Promise applications get accepted at higher rates because the bar is lower? The answer, based on what assessors actually evaluate, is no. The Promise bar is calibrated for under-5-years-of-experience applicants and is judged on trajectory rather than current achievement. The rejection rate is comparable.
Where Talent and Promise diverge meaningfully is the time-to-settlement: Talent grants a 3-year ILR pathway, Promise a 5-year. That difference shows up in the cumulative cost calculation, not in approval probability.
Why the endorsing bodies don't publish this themselves
Two operational reasons. First, the endorser's mandate is to assess against criteria, not to publish performance data. Second, publishing per-endorser rates would invite gaming — applicants choosing the body with the highest pass rate rather than the body their evidence actually fits. The Home Office prefers applicants self-route by discipline, not by approval probability.
The result: the only window into actual endorsement outcomes is FOI requests, practitioner experience and forum aggregation. None of those are clean datasets.
What the data tells you about your application
Three takeaways:
1. First-attempt failure is the modal outcome. If you are mentally budgeting on a 90 percent chance of being endorsed first time, recalibrate. Roughly two of every three first attempts fail at endorsement. That is not because the route is rigged. It is because the bar is calibrated against international evidence standards and most applicants underestimate what "exceptional" means in that context.
2. The visa stage is not the risk. If your endorsement comes through, the visa is highly likely to follow. Plan accordingly: invest disproportionately in the endorsement evidence pack, not in worrying about Stage 2.
3. Evidence quality, not eligibility, is what separates the approved 25–30 percent from the rejected 70–75 percent. The vast majority of rejected applicants are technically eligible. They simply did not present their record in a form that aligned with how assessors read evidence packs. This is where preparation matters most.
How to maximise your success
The single highest-leverage action is honest pre-submission triage. Tech Nation offers a free Endorsement Review for digital-tech applicants — but that review only tells you what's missing, not how to fix it. The fixing is on you.
For applicants who want a structured read on probability before committing 3 to 6 months of preparation work:
- Take the eligibility quiz to get a calibrated profile assessment in under 5 minutes.
- If the result indicates a viable case, book a strategy call. We will walk through what your evidence pack should contain and where the gaps are against current criteria.
- For applicants who already drafted an evidence pack and want a second pair of eyes before submitting, we run a Rejection-Risk Audit that scores the pack against the same dimensions assessors use.
Frequently asked questions
What is the UK Global Talent Visa success rate? Roughly 25 to 30 percent of endorsement applications succeed on first attempt. Once endorsed, well over 90 percent of visa-stage applications are approved. The endorsement is the gate.
Does the Global Talent Visa really have 85–93 percent approval? That figure is the Stage 2 visa-stage approval rate, calculated from the cohort that already has endorsement. It does not include the 70 to 75 percent of applicants who fail at the endorsement stage.
Which discipline has the highest success rate? UKRI-routed research applications appear to clear endorsement at slightly higher rates than digital technology, based on practitioner observation. Volume is much smaller, so single-year figures move sharply.
Has the success rate dropped since January 2025? The bar tightened with the standardisation update. Anecdotally, first-attempt rejections rose for borderline digital-tech cases. Aggregate FOI data for 2025 is not yet public.
Does using a consultant or lawyer change the success rate? There is no FOI split between self-prepared and professionally-supported applications. We believe — and the volume of post-rejection consultations we receive supports this — that professionally-prepared evidence packs reduce variance and avoid the most common drafting errors. We will not invent a percentage we cannot source.
Source: gov.uk / Global Talent visa, transparency data on points-based system grants.