When a client's qualifying income crosses £30,000, the first real decision on their file is which piece of software they keep their records in, because from that point every quarterly update, every year-end and every awkward question in March runs through it. I make that choice from HMRC's recognised list, and this month I found a notice on it saying the list has stopped growing.
Disclosure up front: I work at Xero. Read this as field notes from someone who uses the platform daily, not a neutral review. Xero sits on the recognised list, which makes my employer one of the companies this freeze protects. I have written the rest of this with that in mind rather than around it.
TL;DR: HMRC has stopped letting new software onto the MTD for Income Tax list for 2026-27, and will not publish the replacement rules until early 2027. I think the reason is sound and the timing is bad. Treat today's list as the list, and pick your client's software now.
What has HMRC actually stopped doing?
It has closed the door to new entrants. HMRC's own developer guidance now carries the line that it "is no longer accepting production credential access requests for new 2026-27 quarterly update products, as the market window for these products has now closed" (HMRC Developer Hub, page last updated 15 September 2026).
Production credentials are the keys. Without them a product cannot reach HMRC's live systems, so it cannot file for a real taxpayer at all. An HMRC spokesperson put the reasoning more plainly, saying HMRC had "temporarily paused our clearance of these requests while we revise our access model, to ensure that the MTD products we include on our software list meet the required standards, and provide a good quality experience for customers" (AccountingWEB, 2026).
The pause carries limited exceptions, mainly for applications already in the queue and for existing developers who need access to finish year-end obligations. A replacement approach is promised by early 2027.
Why would HMRC shut the door on new software?
Because a recognised list that admits products which cannot finish the job is worse than a shorter one. This is the part I find genuinely persuasive, and I would rather say so than treat every HMRC decision as an imposition.
The number that makes the case is not the length of the list but how much of it has been proven. The list stood at 165 options as at 4 October 2026, and of those, only 136 had successfully submitted a 2026-27 quarterly update by the 31 August deadline, a figure obtained under a Freedom of Information request (AccountingWEB, 2026). Roughly thirty products were sitting on a recognition badge without a single live filing behind them.
Recognition is what a landlord reads as reassurance. Put a client on a product that has never completed a real submission and you find out in the week of the deadline. I have no appetite for discovering that on someone else's file.
What does a frozen list do to the market?
It raises a barrier to entry, which is a rule that stops new competitors reaching a market rather than an absence of customers wanting to buy. The demand side here is about to move sharply: HMRC expects roughly 1,077,000 additional sole traders and landlords to join from April 2027 (GOV.UK, 5 October 2026).
So the supply of approved products is fixed at the exact moment demand climbs by a million. Everyone already on the list, Xero included, picks up that cohort without a single new rival appearing beside them. Sage, QuickBooks and FreeAgent are in the same protected position. Nobody designed that as policy and I do not think HMRC acted in bad faith, but it is the mechanism, and the companies holding pricing power over the April 2027 cohort are the ones who happened to be inside the door when it shut.
I would rather name that plainly while I work for one of the beneficiaries than discover later that I had been tactful about it.
The 165 options are not 165 real choices
Count the ones you would actually put a client on and the number drops hard. A good portion of the list is thin single-purpose bridging tools, built to push a quarterly figure at HMRC and nothing else. Not every recognised product handles both the quarterly updates and the year-end final declaration that replaces the Self Assessment return, and a product that does one but not the other leaves you assembling the year-end somewhere else.
HMRC's agent toolkit still tells practitioners there is "a wide range of software choices to suit different needs and budgets" (GOV.UK, page last updated 7 July 2026). Both statements are true at once, which is the awkward bit. The range is wide and it is now also fixed.
ICAEW has urged the government to review the first full reporting cycle before the regime extends in April 2027 (AccountingWEB, 2026). Whatever comes of that, it will not add products to the list in time for the clients landing next April.
When do you actually need to have picked?
Earlier than the date in your head. The mandate date is 6 April 2027 for anyone whose qualifying income from self-employment and property was more than £30,000 in the 2025 to 2026 tax year (GOV.UK agent toolkit, 2026). Qualifying income is turnover before expenses, not profit, which still catches people out.
| When | What happens | What I would be doing |
|---|---|---|
| Now to Dec 2026 | List is frozen at its current size | Shortlist two products and test each on a real client file |
| Early 2027 | HMRC publishes its new access model | Read it, but do not wait for it |
| 6 April 2027 | The £30,000 cohort is mandated | Clients already on software, already filing |
The reason I would not leave this to February is that a demo tells you nothing. You learn whether a product copes with a client who has two rental properties and a bit of self-employment by putting that client in it. Start now and you get a full quarter to find out, with the rest of the list still available if the answer is no.
Frequently Asked Questions
Can I use software that is not on HMRC's recognised list for MTD for Income Tax?
No. Submissions go through HMRC's live API, and a product without production credentials cannot reach it. If a provider is not recognised you cannot file a quarterly update or a final declaration through it, however good the bookkeeping side is.
What is a barrier to entry, and why does it matter who is on an approved list?
A barrier to entry is anything that stops a new competitor reaching customers, as opposed to customers not wanting the product. When entry closes and demand rises, the firms already inside face less competition on price and less pressure to improve, without anyone deciding that should happen.
If my client joins MTD in April 2027, when should I actually choose their software?
Before the end of 2026, so you can run a real client file through a shortlist rather than a demo account. HMRC's new access model is not due until early 2027, which falls after the point most practitioners need to have committed.
Does every product on the list handle the year-end as well as the quarterly updates?
No, and this is the check worth making first. Some recognised products cover quarterly updates only and leave the final declaration, which replaces the Self Assessment return, to be done elsewhere. Ask the provider directly which of the two it supports.
What does MTD-compatible software actually cost?
It spans a wide range, which is part of why the choice is worth making early. At the paid end, Xero, my employer, publishes an entry plan at £18 a month after the introductory period with MTD for Income Tax included (Xero, 2026), and the other major providers publish their own rates. The cost worth modelling is the second one: a quarterly-only product means buying something else for the final declaration.
If you have clients heading into the April 2027 cohort and you want a second opinion on the shortlist before you commit a file to it, get in touch. I am happy to talk through what each product actually does at year-end, rather than what the comparison pages say. For the background, the firm's explainers on what changes in April 2026 and what MTD record keeping requires of sole traders cover the basics, and I have written separately on why nobody is backing up your cloud ledger, which is the other question I would ask before putting a client on anything.