The tribunal believed every witness. It accepted that the taxpayer ran a coordinated, genuine project through the worst months of the pandemic, and it said so in terms most judgments reserve for compliments: diligent, innovative, proactive. Then it dismissed the appeal and took the entire £880,286 enhanced deduction with it.

Tanglewood Care Services Ltd v HMRC ([2026] UKFTT 1137 (TC), released 6 August 2026) is a reported First-tier decision squarely on a Covid-operations R&D claim, and it deserves a careful read rather than a headline. Pandemic-era claims built on reasoning similar to Tanglewood's are widely reported to be a large cohort, many enquiries opened into them are still unresolved, and the decision now gives HMRC a reported authority aligned with the position practitioners describe meeting in enquiry correspondence. It also, less noticed, leaves three doors open that a well-evidenced claim could still walk through.

Here is what the tribunal actually decided, and what it means for anyone holding, defending or reviewing a claim of this shape.

Tanglewood operates residential nursing care homes, seven of them in the relevant period, housing predominantly elderly and vulnerable residents. Its original corporation tax return for the year to 31 January 2021, filed in February 2022, contained no R&D claim. An amended return followed in September 2022, adding a single project: "Managing outbreak & effects of a Global Pandemic in a residential care home (Covid-19)".

The activities behind the claim will be familiar to anyone who ran a care setting through 2020: PPE regimes, testing, visitor restrictions, cohorting and isolation, staff segregation, enhanced cleaning, revised admissions, daily reporting. Some measures were adopted ahead of later government guidance. Three witnesses gave evidence, the founder among them, and the tribunal found all three honest and conscientious.

HMRC closed its enquiry by removing the enhanced deduction in full. The dispute that reached the tribunal turned on one question: was any of this "qualifying Chapter 2 expenditure" under the SME scheme, which pulls in the definition of R&D from the BEIS Guidelines, in their 2010 version for a period of this vintage.

HMRC leaned on the absence of a formal, contemporaneous project plan. The tribunal declined to make that fatal: paragraph 19 of the Guidelines does not demand documentation, and a coordinated programme of activities directed at a goal is a project (para [87]). For claimants everywhere, that is a useful holding, because practitioners commonly report meeting a documentation-first line in enquiry correspondence, and HMRC's argument here ran close to it: that paragraph 19 requires a method or plan.

Buried in the reasoning are three findings more taxpayer-friendly than the outcome suggests:

• Capability counts. An advance does not require new underlying scientific knowledge in every case; advances in capability qualify (para [94]).
• Known components do not disqualify. A claim does not fail "simply because the individual measures relied upon by the Appellant were already known" (para [98]).
• System uncertainty is real. In an appropriate case, uncertainty about "how known measures should be combined to achieve a particular outcome may constitute scientific or technological uncertainty" (para [104]).

The decision stops well short of holding that pandemic operations can never be R&D. Its holding is narrower, and sharper: this claim, on this evidence, fell outside the definition.

The core of the case sits in paragraph [99]. The evidence did not establish that the project "sought to achieve an advance in overall knowledge or capability in a field of science or technology". The measures were aimed at the company's own homes, its own residents, its own operations. The Guidelines measure an advance against the field, and an improvement to your own business, however hard-won, points the wrong way.

The tribunal accepted the problems were "real and difficult", then characterised them: "predominantly operational and managerial" (para [106]). Paragraph [108] draws the line that will be quoted in enquiry letters for years: working out "how best to respond to evolving scientific knowledge, public-health advice and practical experience" is "not the same thing as resolving a scientific or technological uncertainty within the meaning of the Guidelines".

The third failure was evidential, and it is the most practical lesson in the decision. No witness held expertise in virology, epidemiology or infection science. Without a competent professional in the relevant field, the tribunal could not assess what the baseline knowledge was, whether the uncertainties were readily deducible, or whether anything advanced the field (paras [112]–[114], following Flame Tree Publishing and AHK Recruitment).

The pandemic produced a wave of R&D claims reasoned much like Tanglewood's: unprecedented problems, solved energetically, written up afterwards. For claims amended into 2022 returns, the window for opening fresh enquiries has passed; the live exposure sits in enquiries already opened and still unresolved, and in discovery assessments where carelessness can be shown, with the extended six-year window for a period ended in early 2021 running into 2027. The department now holds a reported decision aligned with its standing position, and a file built on "nobody had faced this before" and little else has just become measurably harder to defend.

Read against the three open doors, the decision sketches its own counterfactual. A claim that survives looks like this: the advance is framed at field level, not firm level; the uncertainty is specified as a scientific or technological question, ideally a system-integration one; and a competent professional in the actual field, not a director however capable, gives evidence on baseline and deducibility.

Reported decisions do their real work in correspondence, and it is worth predicting the letters now.

Enquiry letters into pandemic-era claims will lean on four passages. Paragraph [99], for the advance pointing inward at the claimant's own operations. Paragraph [106], for uncertainties "predominantly operational and managerial in character". Paragraph [107], where the tribunal reached for the Guidelines' social-science exclusion to characterise the issues as organisation, management and practical operation. Paragraph [117], for the reminder that acting "effectively, innovatively or successfully" is not the statutory test. An adviser who has read the whole decision can meet each quotation with its counterweight from the same judgment, which is a considerably stronger position than meeting it with indignation.

Against [99], paragraph [94]: capability advances qualify, and new underlying science is not required in every case. Against a "known measures" attack, paragraph [98]: known components do not, by themselves, sink a claim. Against "this is just operations", paragraph [104]: uncertainty about combining known measures into a system can be scientific or technological uncertainty. None of these rescues a file without evidence, and all of them reframe what the evidence needs to show.

For weaker files, the decision changes the arithmetic of holding out. A closure notice defended to hearing now runs into a recent, directly analogous authority, with the added problem that the competent-professional gap grows harder to close years after the event: an expert instructed late struggles to anchor a baseline analysis in a record that shows no scientific question being asked. Interest continues to run while positions harden. The honest triage, run early, costs a morning; the alternative can cost the enhanced deduction, the defence fees and the relationship with the client who was promised the claim was safe.

Cases like Tanglewood turn on definitions applied precisely: "advance", "field", "uncertainty", each carrying meaning fixed by the Guidelines and a growing body of FTT authority. This is terrain where a confident summary is dangerous and a cited answer earns its keep. The distinction between paragraph [99] and paragraph [104] of one August decision can be the difference between abandoning a claim and re-scoping it.

That precision is the standard we build AI tax research against: answers grounded in the primary materials, with the source one click away, measured openly on our accuracy benchmark and maintained under a published update policy. For the framework we recommend when evaluating any research tool against fast-moving case law, see our guide to choosing AI tax research software.

Tanglewood lost £880,286 of enhanced deduction while winning the tribunal's respect, which is precisely what makes the decision instructive. The judgment separates four things the pandemic blurred together: effort, novelty, operational excellence and R&D. Only the last one is relieved, and it is proved with field-level advances, specified uncertainties and competent professionals, not with the honest recollections of people who worked heroically through a crisis.

If R&D boundary questions cross your desk, create a GAIN Tax account and get cited answers against the Guidelines and current authority in minutes. More case analysis lives on the GAIN Tax blog.

What was decided in Tanglewood v HMRC? The First-tier Tribunal dismissed the care group's appeal against a closure notice removing £880,286 of enhanced R&D expenditure for the period ended 31 January 2021. The pandemic-management programme was a genuine project but sought no advance in science or technology, and its uncertainties were operational rather than scientific.

Does the decision mean Covid-era claims are all invalid? No. The tribunal expressly held that advances in capability count, known components do not disqualify a claim, and system-integration uncertainty can qualify. The claim failed on characterisation and evidence, and a differently built claim could still succeed.

What is the competent-professional point? Without evidence from someone expert in the relevant field, a tribunal cannot establish the baseline knowledge or whether the uncertainty was readily deducible. In Tanglewood no witness held expertise in virology, epidemiology, infection science or any comparable discipline, and the tribunal noted the relevant field need not necessarily have been virology or epidemiology. The gap was a significant part of why the claim failed, alongside the findings on advance and uncertainty.

Which version of the Guidelines applied? The BEIS Guidelines in their 2010 version, since the accounting period ended 31 January 2021. One footnote for the careful reader: the decision states at [21] that the later revisions "apply only to accounting periods beginning on or after 1 April 2024 (subject to transitional provisions)". The commencement is in fact earlier, SI 2023/293 applying the 2023 Guidelines to accounting periods beginning on or after 1 April 2023, with 1 April 2024 being the merged-scheme date. Nothing turns on it here, where the 2010 version governed either way, but do not carry the tribunal's date into a live file.

Is Tanglewood being appealed? No appeal had been reported as at 30 August 2026. The window for seeking permission runs 56 days from release, to early October 2026, so the position may still change.

What should firms with similar open claims do now? Re-read the file against paragraphs [99], [106] and [112]–[114]: restate the advance at field level if the facts support it, identify a competent professional, and check whether the contemporaneous record shows scientific investigation. Where the file cannot meet the test, quantify the exposure early. Waiting for the enquiry letter surrenders the timing advantage, since interest runs and the evidential gaps only widen as memories fade and staff move on.