A care home's R&D claim reached a tribunal. It should never have been filed.

In July 2023 HMRC wrote to around 7,500 nursing and care home companies. Directors were told their sector was “being targeted by agents or third parties, who encourage them to make Research & Development (R&D) tax relief claims”, and that HMRC rejects most of the claims it sees from it. Care homes, the letter said, “have received unsolicited phone calls from agents, advising that their general business activities qualify for R&D relief”; many owners told HMRC that after the first call they sent over nothing but their accounts and PAYE records. The typical fee was 15 to 25% of whatever came back. Care homes were the first sector HMRC singled out this way.

The pattern was well documented. In evidence to a House of Lords sub-committee published six months before that letter went out, ICAS said “it is my understanding, from looking at the websites of those that do cold calling—the rogue ones—that they are not members of any of the professional bodies.” The National Audit Office later found that “unaccredited agents aggressively solicited taxpayers to submit claims that challenged the definition”. The Comptroller and Auditor General has qualified HMRC’s accounts over error and fraud in R&D relief every year since 2019-20. Error and fraud peaked at a measured 17.6% of the money claimed in 2021-22; the latest measured figure, for 2023-24, is 6.4%.

On 6 August 2026 the First-tier Tribunal released its decision in Tanglewood Care Services Limited v HMRC. A care operator had claimed £880,286 of enhanced R&D expenditure for infection-control work across seven residential care homes in the year to 31 January 2021, HMRC removed the claim, and the appeal was dismissed — five and a half years after the period ended. No earlier care-sector R&D decision appears in the published tribunal record, so this is the first to be read end to end in public. The decision says nothing about how this particular claim came to be made, and the tribunal criticised nobody. It turns on the statutory tests and on the evidence put in front of it.

What the relief is, and what £880,286 means

R&D tax relief reduces a company’s corporation tax bill where it has run a project seeking an advance in science or technology. Effort does not qualify a project, and neither does commercial novelty.

The figure is not a cheque. £880,286 is enhanced R&D expenditure: qualifying costs uplifted by a set percentage, with the enhanced amount deducted from taxable profits before the tax is worked out. A sum taken off the profit, several steps from any cash. The decision states no tax figure anywhere, and the tribunal never decided how much would have qualified (paragraph 3).

Four things have to be true at once:

• a project — defined work with an objective, rather than innovation in general;
• an advance in science or technology — new knowledge or capability in the field itself, not merely in your own company;
• scientific or technological uncertainty — at the outset, nobody knew whether the aim could be achieved, or how;
• a competent professional — someone qualified in that field who could not readily resolve the uncertainty from what is already known.

Our page on what counts as qualifying R&D takes each in turn.

Where we would have stopped

Our own view, plainly. Infection control in care homes during the pandemic was hard, important, operational work. Nothing in it was an advance in a field of science or technology. That is visible in a first conversation, before a form is filled in or a fee is agreed, and saying so is what a company pays a regulated adviser to do. This claim fails at that conversation, and we would not have filed it.

The tribunal found that the company responded in a “diligent, innovative and proactive” manner (paragraph 116), and found its witnesses honest. Take both at face value. Neither defends the claim. Honest, diligent care operators do not spontaneously conclude that their infection-control procedures are a corporation tax event; someone puts that idea in front of them. Who did so here is not in the decision, and we are not going to guess.

What the tribunal accepted

The parts HMRC lost are the most useful thing in the decision.

On the project test, paragraph 19 of the Guidelines does not require a formally documented plan: a coordinated programme of information gathering, review and implementation towards a defined objective sufficed, expressly unlike Hadee Engineering (paragraph 87).

On the advance test the tribunal declined to read paragraph 6 as requiring a claimant “in every case to demonstrate an advance in underlying scientific or technological knowledge”. Advances may arise from resolving uncertainty affecting capability as well as knowledge, and a project is not excluded merely because its individual components are already known (paragraph 94). Read that with the paragraph before, which went HMRC’s way: paragraph 9 cannot be read in isolation, so “a product, process or service does not qualify merely because it exhibits some improved functionality” (paragraph 93).

The tribunal also rejected HMRC’s submission “insofar as it suggests that the claim must fail simply because the individual measures relied upon by the Appellant were already known”, and held that the company need not have been advancing scientific understanding of the virus itself (paragraph 98). It accepted that the Guidelines are “broad enough to encompass system uncertainty arising from the interaction of multiple measures within a system” (paragraph 104), and repeated the point in closing (paragraph 118).

Why the claim failed anyway

The objective, as the tribunal put it, was to determine how best to deploy, balance and manage infection-control measures within the company’s own care homes. That involved refining a great many measures, but “the evidence does not show that those activities were directed towards achieving an advance in overall knowledge or capability beyond the Appellant’s own operations” (paragraph 99). Nothing in the definition measures an advance against the claimant’s own starting position.

A finding made in the company’s favour shows the point sharply: some measures were introduced before similar approaches appeared in later Government guidance (paragraph 80). Being ahead of the official position is worth something, but it is not an advance in a field of science or technology, because the field is not the company.

The uncertainties went the same way. The tribunal accepted that understanding of Covid-19 remained incomplete throughout much of the period (paragraph 103), then held that “uncertainty within the wider scientific and public-health community does not of itself establish that the Appellant’s activities were directed towards resolving scientific or technological uncertainty” (paragraph 105). Staffing arrangements, visitor policies, admissions, PPE procurement and compliance procedures were real and difficult problems, but “predominantly operational and managerial in character” (paragraph 106). Work in the social sciences sits outside science for this purpose, and the tribunal said so (paragraph 107) — a line drawn in what does not count as R&D.

The evidence the tribunal did not have

Three witnesses appeared, and the tribunal found each “honest, conscientious and experienced”, with significant expertise in the care sector (paragraph 111). The difficulty was scope. None “claimed expertise in virology, epidemiology, infectious disease transmission, infection science or any other scientific or technological discipline capable of assisting the Tribunal as to the relevant state of knowledge or capability in the field” (paragraph 112).

The last clause repays reading twice. The absence mattered “not because we consider the relevant field must necessarily have been virology or epidemiology”. It mattered because it left the tribunal without enough evidence on whether the uncertainties were scientific or technological, whether they were readily deducible, or whether an advance beyond the company’s own operations was sought (paragraph 113). That is an evidential finding, not a ruling on which fields a care operator may claim in, and it settles nothing about who counts as a competent professional.

The quietest failure is the last. The activities were extensive and resource-intensive, but the evidence did not establish “a systematic process of investigation or experimentation directed towards resolving scientific or technological uncertainty” (paragraph 115). Effort and expense do not amount to a method. A method is what the records a claim needs capture, kept while the work happens.

Do First-tier Tribunal decisions change the law?

No. FTT decisions bind only the parties and do not set precedent, so a loss on the facts puts no sector outside the relief, and on the main points of principle the tribunal went the claimant’s way (paragraphs 94, 98, 104 and 118). Set it against the Collins Construction and Stage One verdicts, where HMRC declined to appeal and rewrote its guidance. Those two answer who owns a claim; this one answers whether there was a claim.

The period ended 31 January 2021, so the old SME scheme and the 2010 Guidelines applied (paragraph 21); the current edition puts the same four tests behind the merged R&D expenditure credit and ERIS. What an officer looks for has not changed.

What this asks of a claim being prepared now

Four things, each cheap at the start of a claim and impossible at a hearing.

• Frame the advance against the field, not yourself. The officer’s question is what a competent professional in the field could not already do. Your answer goes on the Additional Information Form, normally the first document read.
• Name the field, then the person who can speak to it — someone whose expertise sits in that field rather than in running the business.
• Separate operational difficulty from technological uncertainty in writing, while the work runs. The two feel identical in a hard year and read very differently three years later, when an HMRC enquiry asks which was which.
• Record the method — hypothesis, test, result, next iteration. That is a systematic investigation on paper, and the substance of how a claim withstands scrutiny.

If someone has approached you about a claim and you cannot tell whether the work is R&D, talk it through with a chartered adviser before anything is filed.

Sources

• Tanglewood Care Services Limited v HMRC — the decision, cited as [2026] UKFTT 1137 (TC) and as 01137 on its cover page. Every paragraph number in this article refers to it.
• Guidelines on the meaning of R&D for tax purposes — paragraphs 6, 9, 13, 14 and 19 on advance, uncertainty and what makes a project; 15A on the social sciences, numbered 15 in the 2010 revision applied here.
• HMRC one-to-many letter: R&D tax relief, care homes — the July 2023 letter, on agents targeting the sector, unsolicited phone calls and fees of 15 to 25%.
• HMRC’s approach to R&D tax reliefs 2023–24 — around 7,500 companies written to, the sectors where R&D is unlikely to be carried out, and 17.6% error and fraud for 2021-22.
• Tax measures to encourage economic growth — National Audit Office, HC 445, 31 January 2024, summary paragraph 18.
• House of Lords Economic Affairs Finance Bill Sub-Committee, HL Paper 137 — 31 January 2023, paragraph 92, recording the ICAS evidence quoted above.
• HMRC annual report and accounts 2024–25 — the Comptroller and Auditor General’s qualification on R&D error and fraud.
• HMRC annual report and accounts 2025–26 — error and fraud measured at 6.4% for 2023-24.

This article describes the rules as they stood at the review date above. The rules change: for the current position, start with our guides or talk to us.