R&D tax relief went, in the space of a few years, from one of the most generous and lightly scrutinised reliefs in the UK tax system to one of HMRC's single biggest compliance priorities. Roughly one in five claims now triggers a check. If your company has claimed, or is thinking about claiming, understanding why this shift happened and what HMRC is actually looking for is essential before you file, not after an enquiry letter arrives.

Why HMRC's approach to R&D relief changed so sharply

For years, R&D tax relief operated on a largely self-certified basis: a company described its qualifying activity, calculated the relief, and HMRC processed the great majority of claims without further question. That approach coincided with an explosion in claim volumes, much of it driven by contingency-fee "R&D boutiques" who took a percentage of the relief secured and had every incentive to submit as many claims, in as generous a form, as possible. HMRC's own analysis eventually concluded that error and fraud in the scheme had reached a level it judged unacceptable, and the response has been a sustained, still-intensifying compliance push: more enquiries, restructured internal teams handling them, and a tightening of the qualifying rules themselves through the merged R&D scheme and the additional restrictions on overseas expenditure.

The practical effect for any company claiming relief now is that a claim which would have sailed through five years ago can attract a detailed compliance check today, particularly where it shares the hallmarks HMRC associates with weak or abusive claims: a generic, templated technical narrative; costs that look disproportionate to the company's actual size and activity; and a claim prepared by an adviser working on a contingency fee with no ongoing relationship with the business.

What HMRC actually tests in an enquiry

An R&D enquiry turns on two separate questions, and claims usually fail on one or the other rather than both. The first is whether the underlying activity meets the statutory definition of qualifying R&D at all: a project seeking an advance in science or technology, involving the resolution of scientific or technological uncertainty that could not readily be resolved by a competent professional working in the field. The second, assuming the activity does qualify, is whether the costs claimed are correctly identified and apportioned to that qualifying activity, staff time genuinely spent on the qualifying uncertainty rather than on routine development, correctly treated subcontractor and externally provided worker costs, and correctly excluded ineligible categories.

HMRC's enquiry letters typically request the technical narrative supporting the claim, evidence of the specific uncertainty encountered and how it was resolved, contemporaneous project records, and a detailed cost breakdown reconciled to the company's accounts. A narrative that reads as marketing copy, "our innovative platform revolutionised the industry", rather than a specific, technical account of what was unknown, what was tried, and why the outcome was not obvious in advance, is the clearest single red flag in any enquiry.

The competent professional trap: the legal test is assessed against what a competent professional already working in the relevant field would have known, not against how novel the work felt to the company doing it. A claim that cannot articulate, specifically, what such a professional would already have known and why the uncertainty went beyond that baseline is highly exposed, however genuinely difficult and expensive the project actually was.

Advance assurance and the 2026 pilot

HMRC has long offered an Advance Assurance service for smaller, first-time claimants, giving a degree of comfort before a claim is submitted. From May 2026, HMRC piloted an expanded advance assurance offering allowing claimants to seek assurance on specific elements of a claim, whether a project meets the R&D definition, the treatment of overseas expenditure, contracted-out R&D status, and eligibility for the PAYE/NIC cap exemption, rather than the claim as a whole. Participation is voluntary and the service does not extend to every claimant or every category of question, but for companies with a genuine, ongoing R&D programme it offers a materially lower-risk route than filing and hoping the claim is not selected for enquiry.

HMRC's internal reorganisation of R&D compliance

HMRC has also restructured how R&D enquiries are handled internally, moving responsibility away from the compliance teams that previously handled the volume of post-pandemic claims and towards its established large and mid-sized business compliance teams. The practical consequence for taxpayers is a shift towards more experienced, technically resourced caseworkers conducting enquiries, which tends to mean more detailed, better-targeted questioning, and less tolerance for generic or templated responses than in the earlier, higher-volume phase of the crackdown.

How these enquiries typically unfold

Case A: The claim that looked like every other claim

A software company engages a contingency-fee R&D specialist who prepares a claim built from a standard template, describing "significant technical challenges in scalability and architecture" without identifying any specific, resolvable uncertainty. HMRC opens an enquiry, and the company's own developers, when asked directly what was actually uncertain at the outset of the project, struggle to identify anything beyond ordinary, well-understood software engineering decisions. Unable to reconstruct a genuine technical narrative after the fact, the company agrees to withdraw the claim rather than litigate a position it cannot properly evidence, avoiding a penalty but losing the relief and incurring interest on the repayment.

Case B: The claim that held up under scrutiny

A manufacturing business claims relief for a genuinely difficult materials engineering project, where the company kept detailed lab notebooks, failure logs and dated technical reports throughout the project precisely because the work was experimental. When HMRC opens an enquiry, the company is able to produce a specific, evidenced account of the uncertainty, what was tried, what failed and why, cross-referenced against contemporaneous records rather than reconstructed narrative. The enquiry closes with the claim substantially accepted, the difference being genuine contemporaneous evidence rather than a stronger story told after the event.

The pattern across both cases: HMRC's enquiries reward companies that can show their technical narrative was true at the time, evidenced by contemporaneous project records, not reconstructed convincingly once the enquiry letter lands. Claims built on generic templates rarely survive close questioning, however much money was genuinely spent.

Responding to an R&D enquiry

The first priority is establishing, honestly, whether the underlying claim is defensible: does the technical narrative identify a genuine, specific uncertainty, and can it be evidenced from records that existed before the enquiry started, not assembled afterwards. Where the claim is sound, the response should focus on giving HMRC the specific, technical detail its templated narrative omitted, engaging the actual technical staff who did the work rather than leaving the response entirely to the original claim preparer. Where the claim is weak, an early, managed withdrawal, disclosing the position proactively rather than waiting for HMRC to unpick it, generally produces a materially better penalty outcome than a contested enquiry that ultimately fails.

Received an R&D tax relief enquiry letter?

Speak to us before you respond. Free, confidential 15-minute call.

LONDON: 020 3827 1447 DERBY: 01332 308655

Related guides in this series

Frequently asked questions

How likely is my R&D claim to be enquired into?

Roughly one in five claims now triggers a compliance check, a dramatic rise from a few years ago, driven by HMRC's own assessment that error and fraud had reached unacceptable levels. Claims from high-volume, contingency-fee advisers with generic narratives are disproportionately targeted.

What does HMRC actually test in an R&D enquiry?

Whether the project meets the statutory R&D definition (a genuine advance resolving uncertainty a competent professional couldn't readily resolve), and whether costs are correctly apportioned to qualifying activity. Generic narratives describing ordinary work as groundbreaking are the most common failure point.

Can I get advance assurance that my R&D claim will be accepted?

HMRC offers Advance Assurance for smaller, first-time claimants, and piloted an expanded service from May 2026 covering specific claim aspects. It is not universally available and does not guarantee immunity from a later enquiry if the facts change or were misrepresented.

What is the "competent professional" trap?

The uncertainty must be assessed against what a competent professional in the field already knew, not how novel the work felt to the claimant. Claims that can't articulate this baseline and show the uncertainty went beyond it are highly vulnerable, regardless of genuine cost and effort.