UK R&D Tax Credit Statistics 2026: The Complete Data-Backed Picture
In 2024, UK businesses claimed GBP 7.6 billion in R&D tax relief, across 65,690 claims, the lowest claim volume in 6 years following HMRC's compliance crackdown. This article sets out what that means, who claimed, where, how much, and what it signals for first-time claimants looking at 2026.
1. Headline numbers 2024
The UK's R&D Tax Credits are one of the largest innovation-incentive programmes in Europe by cash outlay. According to HMRC's Research and Development Tax Credits Statistics bulletin, published annually by HMRC, the 2024 picture can be summarised as follows.
Total claims: 65,690. Total relief: GBP 7.6 billion. Year-on-year change in claims: minus 21.1 percent. Average claim (derived): GBP 115,700. Source: HMRC, Research and Development Tax Credits Statistics 2024, provisional figures.
Two things stand out. First, claim volumes fell sharply, driven almost entirely by SME claimants, rather than RDEC. Second, total relief was essentially flat despite the volume decline, which means average claim value rose materially. The short explanation: the compliance crackdown removed weaker claims, and larger, more substantive claims now make up a higher share of the mix.
The five-year view
The 2024 data is best read against a five-year history. From 2020 to 2022, total claim counts rose steadily and total relief edged up with the SME enhanced-rate regime. The first inflection point was the rate reduction from April 2023, followed by the Additional Information Form mandate in August 2023, and the full merged scheme from April 2024. Each of these moved claim counts down or stabilised them, but total relief held broadly steady, signalling that the reforms filtered out low-value claims rather than deterring substantive R&D activity.
Taken together, the trend confirms a common observation inside the profession: the 2020-2022 figures contained a material tail of thin or speculative claims that the 2023-2024 reforms removed. The residual claim book in 2024 is on average more substantive, better documented, and more defensible against HMRC scrutiny than the 2021 or 2022 book.
First-time claimants reading this data in 2026 should therefore not be anchored to the 2022 peak. The headline numbers are now closer to the long-run sustainable level. The practical implication for a first-time claim is that the bar, both in narrative quality and in cost discipline, is higher than it was three years ago.
Aggregate figures vs headline figures
Readers should distinguish between aggregate figures, such as total relief, and claim-level figures, such as average claim. Aggregates are substantive at the national level but unhelpful for benchmarking a single company; averages are more useful at the claimant level but hide wide within-sector variance. We publish both, but when companies ask where they sit, we use sector averages and within-size-band comparisons, not the national average.
2. By sector: who claims
HMRC breaks R&D claim data down by Standard Industrial Classification (SIC) code. The concentration is striking: three sectors account for roughly two-thirds of claims by volume and by relief value.
Information and Communication: 15,830 claims, GBP 2,235m relief, GBP 141k average. Manufacturing: 13,620 claims, GBP 2,165m relief, GBP 159k average. Professional, Scientific and Technical Activities: 13,240 claims, GBP 1,370m relief, GBP 103k average. Wholesale, Retail and Repairs: 5,480 claims, GBP 385m relief. Construction: 4,120 claims, GBP 265m relief. Admin and Support Services: 3,510 claims, GBP 245m relief. Financial and Insurance: 2,170 claims, GBP 315m relief, GBP 145k average. Agriculture, Forestry and Fishing: 1,490 claims, GBP 95m relief. Transport and Storage: 1,120 claims, GBP 85m relief. Arts, Entertainment and Recreation: 950 claims, GBP 70m relief.
The top three sectors dominate both volume and value. Manufacturing has the highest average claim of the top three, reflecting the capital and staff intensity of production R&D. Information and Communication has the highest claim count, reflecting the sheer volume of software businesses with qualifying technical uncertainty. Financial and Insurance has a notably high average claim driven by a smaller number of large claimants in banking, fintech and insurance technology. Source: HMRC R&D Tax Credits Statistics 2024, SIC breakdown table.
What the sector data signals
The dominance of Information and Communication is partly a reflection of the broad definition of qualifying software R&D under HMRC rules. Any project that seeks to resolve scientific or technological uncertainty in software can qualify, and the UK's software sector is large and active. Manufacturing's high average claim reflects the capital-and-staff intensity of physical-product R&D, where costs accumulate faster against a smaller population of claimants.
Professional, Scientific and Technical Activities is a broad SIC category that captures engineering consultancies, architectural practices, specialist scientific services and similar. Claim values here are lower on average than Information and Communication or Manufacturing, but claim volumes are high, reflecting a large and fragmented claimant base.
The long tail, Wholesale and Retail, Construction, Agriculture, Transport, Arts, is where first-time claimants most often ask whether they qualify at all. The short answer: some do, and the common qualifying activities in each are quite specific. Construction claims often relate to novel structural or materials problems. Agriculture claims often relate to crop, soil or livestock-science problems. Transport claims often relate to logistics-optimisation or vehicle engineering. The framing of the scientific or technological uncertainty is the deciding factor.
3. By company size
HMRC publishes a company-size breakdown, which maps to the old SME versus RDEC split and which the merged scheme rules are steadily replacing. For the 2024 data year the mapping is still broadly useful.
Micro (fewer than 10 employees): 34,115 claims (51.9 percent), GBP 1,780m relief (23.4 percent). Small (10 to 49): 15,720 claims (23.9 percent), GBP 1,790m relief. Medium (50 to 249): 8,910 claims (13.6 percent), GBP 1,550m relief. Large (250+): 6,945 claims (10.6 percent), GBP 2,480m relief (32.6 percent).
Micro and small companies account for three-quarters of claim volume, but less than half of relief value. The data shows what intuition suggests: R&D tax relief is used by many small claimants at low value and a smaller number of large claimants at high value. The average claim at the large tier is roughly ten times that at the micro tier, and the compliance environment affects the smaller tier most heavily. Source: HMRC R&D Tax Credits Statistics 2024, enterprise size table.
Implications by size band
For a micro or small business, the 2024 data is reassuring in one sense and sobering in another. Reassuring, because most UK R&D claimants are small; the assessment process is set up for companies of that size. Sobering, because the reforms have raised the fixed cost of claim preparation, and a small claim can no longer support a thin narrative or a casual cost schedule.
For a medium business, the 2024 data shows a claimant tier that moved least. Average claim value is healthy, claim counts held up better than in the SME tier, and the move to the merged scheme has fewer rate-change implications than for loss-making SMEs. The practical step for medium claimants is often to verify whether Capital Allowances or Land Remediation claims are also being picked up, because these are frequently under-claimed at this tier.
For a large business under RDEC, and now under the merged scheme, the mechanical transition has been modest, because the merged scheme uses the above-the-line credit model that RDEC already used. What has changed is the accounting treatment for SME-to-large transitions, where a group structure or growth through acquisition can push the claimant across scheme boundaries within a single accounting period.
4. By region
Regional concentration of claims is significant. HMRC uses the registered office of the claimant company, which means the data over-represents London and the South East relative to where the R&D activity actually happens. Even so, the picture is meaningful.
London: 16,780 claims, GBP 2,210m relief. South East (excluding London): 9,940 claims, GBP 1,320m relief. North West: 6,820 claims, GBP 760m relief. East of England: 6,340 claims, GBP 790m relief. West Midlands: 5,650 claims, GBP 530m relief. Yorkshire and The Humber: 4,210 claims, GBP 390m relief. South West: 4,180 claims, GBP 420m relief. East Midlands: 3,460 claims, GBP 335m relief. North East: 1,560 claims, GBP 140m relief. Scotland: 4,120 claims, GBP 410m relief. Wales: 1,740 claims, GBP 160m relief. Northern Ireland: 890 claims, GBP 135m relief.
London accounts for around a quarter of claims and relief value. The concentration drops sharply outside the South East and North West. Northern Ireland remains the smallest region by both claim count and relief, reflecting the smaller population of registered innovation-active companies, rather than any policy difference. Scotland and Wales together account for around 9 percent of claims and 7.5 percent of relief. Source: HMRC R&D Tax Credits Statistics 2024, regional breakdown table.
Regional caveats
The regional table has two known biases that readers should hold in mind. First, the registered office bias: a group headquartered in London with R&D activity in Leeds or Cardiff will appear under London in the table. For policy discussion, this overstates the South East. For company-level benchmarking, it has little effect. Second, the submission-date bias: regional tables are compiled from claims received in the statistical year, which lags the activity year; a change in regional activity takes one to two years to show in the data.
Scotland and Wales, taken together, account for a meaningful share of claim counts but a smaller share of relief, consistent with a larger proportion of smaller claimants in both nations. Northern Ireland's smaller share reflects the smaller population of registered R&D-active companies, not any rule difference. Companies across all four UK nations file under the same HMRC rules, with no separate devolved regime for R&D Tax Credits.
5. Merged scheme transition: year-1 data
The merged R&D scheme took effect for accounting periods beginning on or after 1 April 2024. The 2024 statistical release therefore captures only a partial view of claims submitted under the new rules; most of the claim data in the 2024 bulletin relates to accounting periods predating the merger. What the year-1 data does show is the direction of travel.
Three early-signal effects are visible in the 2024 data. First, SME claim volume is concentrated in pre-merger periods. The sharp volume decline is more pronounced for SME claims, consistent with an end-of-era surge of pre-merger filings, followed by a pause as companies and advisers adapt to the new rules. Full first-year merged-scheme data will appear in the 2025 statistics release.
Second, average claim size is rising. The mix shift toward larger, more substantive claims is consistent with both the compliance crackdown and the merged scheme's reduced benefit rate at the SME end, which makes low-value claims marginal on a fee-for-outcome basis.
Third, ERIS uptake has begun. The Enhanced R&D Intensive Support regime, introduced alongside the merged scheme for loss-making companies with at least 30 percent R&D intensity, is starting to appear in the published tables. Uptake is concentrated in biotech, deep-tech and specialist software sub-sectors. For background on the rules themselves, see the merged scheme explained. Source: HMRC R&D Tax Credits Statistics 2024, supplementary tables on scheme type.
ERIS eligibility in practice
The 30 percent R&D intensity threshold for ERIS is a simple arithmetic test but a demanding one in practice. It is computed across the accounting period: qualifying R&D expenditure divided by total expenditure, where total expenditure uses the definition in the legislation (not accounting profit and loss). For early-stage biotech and deep-tech businesses, where most of the team is working on the core research problem, passing 30 percent is straightforward. For growth-stage software businesses with material sales and marketing spend, it is harder.
A company's ERIS status can change year-on-year. A biotech company in clinical trials may pass the threshold one year and fall below it the next, as commercial spend ramps. The practical consequence is that ERIS cannot be treated as a permanent regime for a given company; it must be assessed per accounting period.
AIF adoption in year one
Adoption of the Additional Information Form, mandatory from 1 August 2023, has been essentially universal among HMRC-registered specialist firms. The bottleneck has been the internal drafting time: a good AIF requires the competent professional to write or co-write the project narrative, which is time the internal team may not have budgeted for. Specialist firms that run structured interviews with the competent professional, and draft the AIF on that basis, have absorbed most of this cost on behalf of the claimant.
6. Compliance landscape: the enquiry rate
The compliance environment for R&D claims tightened sharply from 2022 onward. HMRC's Fraud Investigation Service and the dedicated R&D compliance teams have grown; the enquiry rate has risen materially; and the Additional Information Form, mandatory since August 2023, has filtered out a number of claims that previously reached the system without structured supporting evidence.
HMRC's published estimate of the level of error and fraud in R&D claims in recent years has been in the range of 16 to 20 percent of total claim value. The enquiry rate observed by advisers operating in the market has been broadly consistent with that, at approximately 20 percent of all claims open to some level of HMRC review.
The most common HMRC enquiry triggers, based on published HMRC guidance and specialist adviser reports, are these. 1. Thin technical narrative: project descriptions that do not clearly identify scientific or technological uncertainty or the advance sought. 2. Generic AIF: Additional Information Forms that appear boilerplate or copy-pasted across projects. 3. Software claims without algorithmic specificity: software R&D descriptions that restate the product, rather than the underlying technical advance. 4. Subcontractor anomalies: large, round-number subcontractor costs, especially to connected parties or overseas entities. 5. Out-of-scope expenditure: sales, marketing, admin or general training costs included in qualifying expenditure. 6. Full-year apportionment of staff time: 100 percent of one or more staff treated as R&D, particularly for senior or generalist roles. 7. Inconsistent year-on-year claims: a sudden spike in claim value with no corresponding change in company activity. 8. Failure under the competent professional test: narratives that do not demonstrate that the competent professional was unable to resolve the uncertainty. 9. Grant interaction errors: Innovate UK or other grant-funded projects claimed incorrectly under the wrong regime. 10. Cost allocation errors: capital costs or unrelated operating costs miscategorised as qualifying R&D expenditure.
An enquiry does not mean a rejected claim. A well-prepared claim defended by a specialist adviser frequently stands. What it does mean is more time, more documentation, and more stress for the client. The practical response is to prepare the claim to enquiry standard from the outset. Source: HMRC R&D compliance action updates, HMRC error-and-fraud estimates, and specialist adviser reports published in trade press 2023 to 2025.
Enquiry outcomes
Of claims that enter enquiry, a substantial majority are resolved without the claim being rejected outright, particularly where a specialist adviser engages early and provides structured responses. Full rejection of a claim is relatively uncommon when the underlying R&D is real and the documentation is in order; partial rejection, where specific cost categories or project elements are removed, is more common. The time cost of an enquiry, and the management attention required, is the main practical burden on the claimant.
A well-prepared claim is not automatically exempt from enquiry. HMRC runs random enquiries as part of its compliance framework. A specialist adviser's job in an enquiry is threefold: respond promptly and substantively to HMRC's questions, hold the line on defensible positions, and concede points that are not defensible rather than extending the process.
7. Why claim volumes fell 21 percent post-2023
The 21 percent decline in SME claim volume between the 2023 and 2024 bulletins is material and has several contributing causes. Three stand out.
Rate reduction for SMEs from April 2023. Pre-April 2023, the loss-making SME payable credit produced approximately 33 pence per GBP 1 of qualifying spend at peak. From April 2023 this was reduced materially for non-intensive loss-making SMEs, with ERIS introduced to protect the most R&D-intensive. For marginal claims on a fee-for-outcome basis, the rate change made preparation uneconomic, and a number of claims that would previously have been filed stopped being filed.
Mandatory Additional Information Form from August 2023. The AIF required structured project descriptions, cost breakdowns, a responsible officer at the company, and adviser details. For claims that had been prepared thinly in the past, the AIF either required meaningful additional work or exposed thinness that would not survive HMRC scrutiny. A number of smaller advisers exited the market in 2023 and 2024.
Compliance crackdown. HMRC's expanded R&D compliance teams, together with the error-and-fraud estimates published in 2022 and 2023, changed the risk-reward profile for speculative claims. Claims that would historically have been submitted, and paid with little friction, now face a realistic probability of enquiry. Some claimants chose not to claim.
The effect is concentrated at the lower end of the claim size distribution. Larger, well-prepared claims have been less affected. The overall decline in relief (approximately 1.3 percent) is much smaller than the decline in claim count (approximately 21 percent), because the claims that disappeared were mostly smaller and thinner ones. Source: HMRC R&D Tax Credits Statistics 2024 commentary, Spring and Autumn Statement 2022 to 2024 scheme-rate announcements.
The adviser market
The compliance tightening has had a second-order effect on the adviser market. The population of R&D advisers in 2020-2022 was larger and more varied than it is today. A subset of advisers who built a business around high-volume, low-touch claim preparation has either exited, merged with larger firms, or substantially changed approach. The remaining adviser market is more concentrated around firms with structured processes, documented compliance postures and qualified tax advisers at firm level.
This consolidation has three consequences for claimants. Fees are more uniform than they were. Quality variance within the market is narrower. The time required from the claimant's competent professional has increased on average, because the remaining firms do not cut corners on the AIF.
8. Average claim value by industry
For companies benchmarking their own claim against comparable businesses, the sector-average claim size is a useful reference point. Average claim size by top-10 sector for 2024:
Manufacturing GBP 159k. Financial and Insurance GBP 145k. Information and Communication GBP 141k. Professional, Scientific and Technical GBP 103k. Transport and Storage GBP 76k. Arts, Entertainment and Recreation GBP 74k. Wholesale, Retail and Repairs GBP 70k. Admin and Support Services GBP 70k. Construction GBP 64k. Agriculture, Forestry and Fishing GBP 64k.
The headline: Manufacturing, Information and Communication, and Financial and Insurance sit at the top of the average-claim table. These are the sectors where qualifying R&D activity tends to be intensive, and where the workforce contributing to R&D is expensive. Construction and Agriculture sit at the bottom, but this does not mean claims are not worth preparing in those sectors. It means the typical claim is smaller, and the eligibility analysis often matters more, than in sectors where a claim profile is obvious. Source: HMRC R&D Tax Credits Statistics 2024, sector and SIC breakdown, average claim values derived.
Benchmarking caveats
Sector averages are a starting point, not an endpoint. A software-as-a-service business and a medical-device firm both sit under Information and Communication or Professional, Scientific and Technical in the SIC classification, but their claim profiles look very different. A SaaS business claim is typically staff-heavy with modest consumables; a medical-device firm's claim is often staff-plus-consumables with a significant regulatory-testing element. Averaging across these sub-sectors loses information.
For a rigorous benchmark, specialists work with sub-sector data where available and with matched-peer samples from prior engagements. That said, the headline average for a sector is a useful sanity check. If your draft claim is an order of magnitude different from the sector average, absent a clear reason, that is a signal to re-examine the cost schedule before filing.
Seasonality in claim submission
Claim submissions are seasonal. The largest volume of claims is filed in the six weeks before the two-year post-period-end deadline, concentrated in the months following common year-ends (March, April, September and December in the UK). HMRC's processing times for R&D claims are themselves variable and can lengthen around these peaks. For companies that can file earlier in the window, the processing experience is typically smoother.
9. What this means for first-time claimants in 2026
A first-time R&D claimant looking at the 2026 environment should take away five things from the 2024 data.
One. The scheme is still large, and still active. GBP 7.6 billion of relief in one year is meaningful fiscal support, and HMRC has repeatedly stated a policy intention to continue supporting UK R&D through the merged scheme and ERIS. First-time claimants are not arriving at a closing door.
Two. The bar is higher than it was. The compliance crackdown is real. Claims that would have been paid easily in 2019 or 2020, now face meaningful HMRC scrutiny. A first-time claim in 2026 needs to be prepared to enquiry standard from day one.
Three. Sector matters. Information and Communication, Manufacturing, and Professional, Scientific and Technical sectors dominate. A software, engineering or scientific services company has the benefit of HMRC's deep familiarity with claims in those spaces. A less-represented sector needs clearer framing of the scientific or technological uncertainty.
Four. The intensity test matters for loss-making companies. If your company is loss-making, the 30 percent R&D intensity threshold is the critical line. Above it, ERIS; below it, the merged scheme. The difference in cash value is material.
Five. Capital Allowances and Land Remediation are often larger than R&D. A first-time claimant who has just fitted out premises or acquired contaminated land may find the Capital Allowances or Land Remediation opportunity exceeds the R&D claim. This is the reason the Uplift Tax assessment looks at all three together.
For first-time claimants, the single most useful step is an independent eligibility view before committing to a preparer. Our process page describes how that works.
What to ask before engaging a preparer
From the 2024 data, five questions are worth asking any adviser before engaging them. Does the firm hold CTA or ATT qualifications at firm level, and is it on the HMRC agent register? How does the firm structure AIF drafting: interview-led, client-led, or hybrid? What is the firm's enquiry rate over the last two years, and how are enquiries resolved? Is the fee no-win-no-fee and disclosed in writing before engagement? Will the fee cover HMRC enquiry support, or is that a separate line?
These five questions filter out most issues at the front door. A firm that cannot answer them directly is not the right firm for a first-time claim in 2026. A firm that answers them clearly is worth continuing the conversation with.
10. Methodology and sources
All figures on this page are drawn from HMRC's Research and Development Tax Credits Statistics annual bulletin (the 2024 release), with supplementary references to HMRC's error-and-fraud estimates, HMRC compliance updates, Spring and Autumn Statement announcements, and specialist adviser trade press published between 2023 and 2026. Where an average or derived figure is shown, we have noted it as derived; the underlying totals come directly from the HMRC tables.
Key assumptions. "2024" refers to HMRC's 2023-24 data year, as published in the 2024 statistics bulletin. These are provisional figures at the time of publication. Sector breakdowns use HMRC's SIC-level aggregations. For claimants whose SIC is not assigned in the return, HMRC uses a residual category; these are excluded from sector tables. Regional data is based on the registered office of the claimant company, which can differ from the location of the R&D activity. Company-size categories use HMRC's enterprise size fields. For companies undergoing reclassification (for example, moving from SME to large), a single period is assigned to one category. Final HMRC figures typically revise provisional figures by a small amount in subsequent annual releases; the direction of the revision is usually upward for claim count and relief.
Primary sources: HMRC, Research and Development Tax Credits Statistics, annual bulletin, 2024 release (gov.uk). HMRC, Measuring Tax Gaps, 2023 and 2024 editions (gov.uk). HMRC, Additional Information Form guidance, published August 2023 and updated periodically (gov.uk). HM Treasury, Spring and Autumn Statement documents 2022, 2023, 2024 (gov.uk). Specialist trade press coverage of R&D compliance activity, 2023 to 2026.
Dates pulled: the 2024 HMRC R&D Tax Credits Statistics release was reviewed at the time of writing (19 April 2026). Figures are subject to HMRC revision.
Use this data
These statistics are republished from HMRC's R&D Tax Credit Statistics 2024 release. You may quote figures from this page in articles, presentations, internal reports, academic work or social media with attribution to Uplift Tax (uplifttax.com), and the underlying HMRC source. Please link back to this page where practical, so readers can check the methodology.
If you would like to embed a light summary of the headline numbers in your own site, the following iframe embed is available for journalists, bloggers and industry analysts: <iframe src="https://www.uplifttax.com/embed/stats-2026/" width="100%" height="620" frameborder="0" title="UK R&D Tax Credit Statistics 2026: Uplift Tax"></iframe>. The embed is a self-contained widget showing headline figures, and updates automatically when we refresh the data.
If you have questions on the underlying figures, or would like to discuss the data with someone before publishing, please contact us via the contact page.
Frequently Asked Questions
How many R&D tax credit claims did UK companies make in 2024?
According to HMRC's Research and Development Tax Credits Statistics bulletin, UK businesses filed 65,690 R&D claims for the 2023-24 data year, the lowest claim volume in six years, following the 2023 compliance reforms.
How much R&D relief did HMRC pay in 2024?
HMRC paid approximately GBP 7.6 billion in R&D relief across the 65,690 claims reported in 2024, an average of around GBP 115,700 per claim. These are provisional figures at time of writing and are subject to HMRC revision.
Why did R&D claim volumes fall in 2024?
Claim volumes fell approximately 21 percent from the prior year. Three factors: the rate reduction for SMEs from April 2023, the mandatory Additional Information Form from August 2023, and the sustained HMRC compliance crackdown, which raised the enquiry rate and deterred weak or speculative claims.
Which sectors claim the most R&D relief?
Information and Communication (software and technology), Manufacturing, and Professional, Scientific and Technical Activities together account for approximately two-thirds of all UK R&D claims by volume and by value. Manufacturing tends to have the highest average claim value. Information and Communication has the highest claim count.
Where Does Your Company Sit in This Data?
A free Uplift Tax assessment puts your qualifying expenditure next to HMRC sector and size benchmarks, gives you an indicative claim value, and tells you whether R&D, Capital Allowances or Land Remediation is the better-value relief for your situation.
Request Your Free Assessment