HMRC Random R&D Enquiry: What MREP Means If You’re Selected

The Mandatory Random Enquiry Programme, known as MREP, is HMRC's structured approach to measuring error and fraud across the R&D tax credit scheme. HMRC selects a representative sample of filed SME claims at random, examines each one in detail, and uses the results to track overall compliance levels. The programme is described by HMRC as good practice in the measurement of error and fraud. Three MREP rounds have been completed. The most recent published results, MREP2, covered 2021 to 2022 SME claims and found that 49 per cent were fully compliant, 30 per cent were wholly non-compliant, and 21 per cent were partially non-compliant.

Why HMRC Introduced MREP

Before MREP, HMRC's understanding of the overall non-compliance rate in R&D tax credits was incomplete. Compliance activity was largely reactive: HMRC opened enquiries based on risk signals, but had no reliable measure of how much non-compliance existed in claims that did not trigger those signals. This left the overall error and fraud rate as an estimate with wide uncertainty.

HMRC introduced MREP to provide a statistically valid baseline. By selecting claims at random within stratified bands, rather than only those that appeared suspicious, HMRC gets a picture of compliance across the whole population of claimants. The MREP2 results, published in HMRC's Approach to R&D Tax Reliefs 2023 to 2024 report, produced the specific figures that underpinned the government's decision to extend and strengthen the compliance regime.

The headline finding from MREP2 was that the overall error and fraud rate for R&D tax relief in 2021 to 2022 was 17.6 per cent, equivalent to approximately £1.3 billion of the £7.6 billion claimed. The subsequent annual report for 2023 to 2024 showed that rate had fallen to 7.8 per cent, and the most recent report recorded a further fall to 5.9 per cent. HMRC attributes the reduction to the package of reforms introduced since 2023, of which the Additional Information Form, the mandatory agent registration scheme, and MREP itself are part.

How MREP Selection Works

MREP selection is random, but it is random within a structured sample. HMRC divides the population of SME R&D claims into strata, for example by claim size band or by first-time versus experienced claimer status, and selects a representative sample from each stratum. This means both small and large claims, new and established claimers, can be selected. No claim is exempt from selection by virtue of its size, the sector, or the claimant's history.

Being selected does not mean HMRC suspects anything. It means your claim has been drawn in the random sample for that period. The MREP2 data showed that fewer than 10 per cent of the cases examined had any fraud indicators. The majority of non-compliance found was error rather than deliberate overclaiming.

HMRC has 12 months from the filing date of your Corporation Tax return to open a formal compliance check. For standard MREP selections, the check is typically opened within that window. If HMRC suspects deliberate error or fraud, the time limit for opening an enquiry extends to four years (for carelessness) or 20 years (for deliberate conduct).

What Happens When You Are Selected

HMRC will write to the company to notify it that the R&D claim is being reviewed as part of the random enquiry programme. The letter will explain that selection is random and not based on a specific concern about the claim. It will ask the company to provide supporting information and documentation within a stated period.

At this stage, the first thing to do is notify your R&D adviser. The adviser should manage the response, prepare the evidence pack, and handle all HMRC correspondence. If you do not have a specialist R&D adviser currently engaged, now is the time to appoint one. Uplift Tax introduces companies to HMRC-registered specialists who handle enquiry responses as well as claim preparation.

The initial HMRC request

HMRC will typically ask for the detailed information underlying your claim. This means the project-level narratives, the cost schedules, the records that support apportionment of staff time, contracts with subcontractors or externally provided workers, and any other documentation relating to the qualifying expenditure categories you have claimed. The request is specific to the information HMRC needs to verify your position, not a general data dump.

The response window

HMRC will give a deadline for the response. This is typically 30 to 60 days from the date of the enquiry letter. If you need more time, contact HMRC or ask your adviser to do so. HMRC will generally grant a reasonable extension if asked promptly. Failing to respond without making contact is not a position you want to be in.

HMRC review and follow-up questions

Once HMRC has received your response, it will review the evidence. In most MREP cases, HMRC will either accept the claim as filed, propose adjustments to specific cost items or projects, or ask follow-up questions about specific technical or cost matters. A claims officer may call to discuss the technical content. Where your project narratives are detailed and specific, this call is usually straightforward. Where narratives are generic or do not clearly demonstrate the advance and uncertainty tests, it becomes more difficult.

Outcome letter

HMRC will issue a closure notice at the end of the check. This either confirms the claim as filed, or sets out any amendments and the tax implications. If HMRC has found overclaimed amounts, you will owe the excess relief back plus interest. In cases where HMRC finds deliberate inaccuracy, penalties can apply on top. For fully compliant claims, the closure notice ends the process and the refund is released if it was withheld pending the check.

What the Compliance Data Shows

HMRC's MREP2 data, published in the Approach to R&D Tax Reliefs 2023 to 2024, provides the most detailed picture of compliance by claim type.

The compliance gap between large and small claims is striking. For claims above £1 million of qualifying expenditure, the MREP2 compliance rate was 75 per cent. For claims below £1 million, compliance rates ranged from 29 to 58 per cent depending on the size band. First-time claimants had higher non-compliance rates than experienced claimers: 46 per cent of first-time claimers were wholly non-compliant, compared to 24 per cent partially non-compliant among experienced claimers.

HMRC's interpretation is that smaller, first-time claims are disproportionately affected by poor-quality advice. Where a company does not fully understand what R&D means for tax purposes, or where an adviser has included activities that do not qualify, the claim fails. Where a company has sound processes and understands the qualification tests, the claim usually holds up.

Building a Strong Evidence Pack

The best time to prepare for MREP is before you file the claim, not after you receive an enquiry letter. The evidence you need falls into four categories:

Project-level records

HMRC needs to see that the R&D activities actually happened as described. This means contemporaneous records: lab notebooks, sprint retrospectives, design meeting notes, test results, prototype iterations, or research papers. These records do not need to be formatted for an R&D claim. HMRC's assessors are technically trained and can read engineering documents, code repositories, or scientific reports. What matters is that the records are contemporaneous (created at the time of the R&D, not written up afterwards) and specific (showing the actual uncertainties encountered and how they were approached).

Time apportionment evidence

Staff time is usually the largest item in an R&D claim. HMRC expects to see how time was divided between qualifying R&D and business-as-usual activities. Contemporaneous timesheets are the gold standard. Where timesheets were not kept, HMRC will accept reasonable estimates supported by project plans, sprint records, or manager statements, but the methodology must be documented and defensible. See the staffing costs guidance for detail on what HMRC expects.

Cost evidence

For each expenditure category claimed, HMRC will want to see the underlying financial records: payroll records, supplier invoices, contracts, and expense claims. The costs in the claim should tie back to these records. Any apportionment between R&D and non-R&D use needs to be explained and evidenced.

The Additional Information Form narrative

The AIF narratives you submitted are the starting point for HMRC's review. If the narratives are vague or generic, HMRC will ask for more detail. If they are specific and technically accurate, the review is likely to be straightforward. A strong AIF is the best single investment in MREP-readiness you can make.

After the Enquiry: What to Do Next

Whether your MREP check results in a clean bill of health or in adjustments, it is worth reviewing your claim process in light of the experience. If HMRC found adjustments, understand what drove them and build that understanding into your next claim. If the check closed cleanly, make sure your record-keeping and AIF processes are documented so they can be repeated.

MREP is not a one-off exercise. HMRC intends to run further rounds, and the programme is now embedded as an ongoing part of R&D tax relief compliance. The probability of any individual claim being selected in any given year is a function of how many claims are filed and how large the MREP sample is. It is not possible to predict or avoid selection. What is possible is to ensure that if you are selected, your claim will withstand the scrutiny.

To check whether your current R&D claim documentation would withstand an MREP review, get a free assessment through Uplift Tax. HMRC-registered specialists can review your AIF narratives, cost schedules, and records before the claim is filed.