Where to Launch First: EU, UK, or US? A Founder's Guide

Status as of 21 July 2026 (key dates re-verified against MHRA and GOV.UK sources), next review 1 October 2026 or immediately on any material development in the UK's Medical Devices (Amendment) Regulations 2026 process, whichever is sooner.

Founders usually ask this question as if it were about ambition: which market matters most to us. It is actually a question about your device class in each system, and the answer to that question can point you somewhere your ambition wouldn't have picked. It is one of several sequencing decisions covered in where to start with health tech regulation.

In short: There is no universal first market. Sequencing depends on your device class in each system: standalone software that is Class IIa under EU MDR can, today, sometimes remain self-certified Class I in the UK under the legacy framework, making the UK a faster first launch pre-market — though GB post-market surveillance obligations, materially strengthened since June 2025, apply in full either way. The MHRA's draft Medical Devices (Amendment) Regulations 2026, published 8 May 2026, does not close this arbitrage the way many expected — on our reading of the current draft, SaMD would generally remain Class I in Great Britain, whose confirmed software-specific changes are a new international reliance pathway, a mandatory unique device identifier (UDI) requirement, and a predetermined change control plan (PCCP) mechanism, not a reclassification. That means there is currently no confirmed mechanism or date for this arbitrage closing, not a confident mid-2027 estimate. US 510(k) timelines depend heavily on predicate availability.

The question behind the question

"Where should we launch first" usually collapses three different questions into one: where is the revenue, where is the evidence cheapest to generate, and where is the regulatory path fastest. These three answers are frequently different markets, and founders who pick a first market on revenue potential alone sometimes discover, months into a submission, that they picked the slowest and most expensive regulatory path available to them for a device at their current stage of evidence.

The more useful framing is to treat market sequencing as a portfolio decision: which market gets you a CE mark, UKCA mark, or FDA clearance fastest, at a stage where your evidence is thinnest, so that you can generate real-world data and revenue to support the harder markets later. This is not the same question as which market has the biggest addressable opportunity, and the two should be evaluated separately before you commit a roadmap to either.

The three systems compared

All figures above are an industry-cited range, including regulatory consultancies and market analysts such as posos.co, not Venitara quotes, and vary significantly with how mature your technical file and evidence base already are when you start. CE Marking Cost and Timeline for Medical Software breaks the EU figures down by artefact, and Why Most 510(k)s Get an Information Request does the same for the US route.

The UK legacy-framework arbitrage, explained honestly

The single most consequential, and most commonly missed, fact in this comparison is that the UK's post-Brexit medical device framework, under UK MDR 2002 (as amended), retains transitional arrangements that let some devices continue to self-certify under legacy Class I rules even where the equivalent EU MDR classification would require Notified Body involvement. For a standalone software product that MDR's Rule 11 pushes into Class IIa in the EU, this frequently means the same product remains Class I, self-certified, in Great Britain — a real and currently exploitable divergence between the two frameworks, not a loophole or an edge case.

One thing the "faster, cheaper" framing must not obscure: this is a pre-market arbitrage only. The Medical Devices (Post-market Surveillance) Regulations 2024 have applied to all devices on the GB market since 16 June 2025 — including legacy Class I self-certified software — bringing enhanced post-market surveillance, trend reporting, periodic summary reporting, and vigilance obligations. Self-certifying Class I in Great Britain in 2026 lowers the pre-market barrier, but it does not put you in a light-touch regime once you're on the market, and a GB launch budget needs to resource those post-market obligations from day one. Nor is self-certification itself zero work: it still requires MHRA registration, a UK Responsible Person for non-UK manufacturers, technical documentation, and a documented conformity position.

Worth being clear about what the three confirmed software-specific changes actually mean in practice, since they matter even though they don't touch the Class I question directly. The international reliance pathway would let a device already authorized in Australia, Canada, or the US reach the GB market through that existing authorization rather than a fresh UK assessment; the MHRA has flagged this as landing around 2028, well after the Draft Amendment's own anticipated in-force date. For a founder holding US 510(k) clearance or planning to, this is a second route into GB that doesn't exist today, arriving on a separate and later timeline than the rest of the Draft Amendment. The mandatory UDI requirement brings GB into closer alignment with how the EU and US already track devices post-market, which is largely an operational and labelling cost rather than a classification change. The PCCP mechanism for SaMD would let a manufacturer pre-specify the bounds of future software modifications, similar in spirit to FDA's own PCCP guidance for AI-enabled device software, so that qualifying updates don't each require a fresh regulatory touchpoint. None of these three changes reclassifies anything; they add routes and mechanisms around the classification question, which is exactly why the Class I self-certification path for standalone software is, on the current draft, still there.

That said, "not confirmed to be closing under this specific draft" is not the same as "guaranteed to remain open indefinitely." The UK has moved faster and less predictably on medical device regulation than the EU has in the years since Brexit, and a separate, earlier MHRA consultation on indefinite recognition of CE-marked devices shows the underlying policy questions here are still actively being worked out even outside this specific Draft Amendment. A founder building a multi-year roadmap around this arbitrage should treat its long-term durability as genuinely uncertain, not because of anything specific in the current draft, but because of the general pace and unpredictability of UK regulatory reform since 2021. That uncertainty is precisely why this page carries a dated status line rather than a one-time verdict.

A separate current option worth knowing about: Great Britain still accepts CE-marked devices on a transitional basis, with end dates that depend on device type. Under the current arrangements (set by the Medical Devices (Amendment) (Great Britain) Regulations 2023), devices CE-marked under the old EU directives (MDD or AIMDD) are accepted until the sooner of certificate expiry or 30 June 2028; IVDs under the IVDD until the sooner of certificate expiry or 30 June 2030; and devices compliant with the current EU MDR or IVDR until 30 June 2030. A founder who already holds — or will soon hold — a CE mark may therefore be able to place the CE-marked device on the GB market without a separate UKCA process while that acceptance lasts — and possibly for longer: an MHRA consultation that closed in April 2026 proposed recognising EU MDR and IVDR CE marks in Great Britain indefinitely, and extending the MDD date to 31 December 2028. That is a proposal, not law, so plan against the dates above and check the current position for your device type before building on it.

A worked example. Consider a Class IIa-equivalent clinical decision-support product, still pre-revenue, with a design history file and verification testing complete but no clinical evaluation report yet drafted. Under EU MDR, this product needs Notified Body involvement before it can carry a CE mark, a process that, per the industry-cited range above, commonly runs 12 to 18 months once formally underway. Under the current UK legacy framework, if the product's classification history and current transitional guidance support it, the same product can potentially self-certify as Class I in Great Britain: no Notified Body review, a materially shorter internal timeline, and a launch that generates real users, real performance data, and real revenue while the EU submission is still being assembled. That UK launch does not substitute for the EU MDR work, and it does not shortcut it, but it can convert a 12-to-18-month gap with no market presence into a period where the product is already live somewhere, earning the clinical and commercial evidence that strengthens the EU file when it's eventually submitted. That GB launch still comes with MHRA registration, a UK Responsible Person, and the full GB post-market surveillance obligations described above — faster to market, not free of obligations.

When US-first makes sense

The UK arbitrage does not mean the UK is always the right first market. US-first makes more sense in a specific set of situations: when a clear predicate device already exists for a straightforward 510(k) comparison, since predicate availability is the single biggest driver of US timeline and cost; when your primary investors, advisors, or initial customer base are US-based and a US regulatory milestone is what moves your next funding round; or when your device's clinical evidence strategy is easier to build against US clinical trial infrastructure and patient populations than European equivalents. One US-side update to budget for either way: FDA's Quality Management System Regulation (QMSR) has been in force since 2 February 2026, incorporating ISO 13485:2016 by reference, so a US launch now carries an ISO 13485-shaped quality-system expectation from the start.

US-first is a harder case when no clean predicate exists, since a De Novo submission takes longer and requires more original evidence, or when your total addressable market is genuinely EU-weighted and a US-first sequence delays revenue in your actual core market for a regulatory milestone with limited immediate commercial value there.

Sequencing playbooks for three founder archetypes

The evidence-thin early-stage team, pre-revenue or early-revenue, with a Class IIa-equivalent software product and thin real-world evidence, is often best served by capturing the UK legacy arbitrage where it currently applies, using that faster, cheaper first launch to generate real-world performance data, then using that data to strengthen the EU MDR submission that was always coming. This sequence trades a lower initial addressable market for meaningfully de-risked and better-evidenced later submissions.

The US-anchored team, with US-based investors, a US-first customer pipeline, and a clean 510(k) predicate available, is often best served by a US-first sequence, using FDA clearance as the milestone that unlocks the next funding round and US commercial traction, before investing in the parallel EU MDR build. CE Mark vs FDA Clearance: The Real Differences, read in reverse, is equally relevant here: your US evidence will feed the EU file later, even though the two submissions remain legally separate.

The EU-core team, with a clinical and commercial base concentrated in DACH or wider EU markets, is often best served by building the MDR technical file as the primary track from day one, potentially using a UK legacy launch opportunistically alongside it rather than instead of it, since the EU MDR work was always going to be the larger and more central build regardless of sequencing.

None of these three archetypes is a guarantee of a specific outcome or timeline; they are patterns we see repeatedly, and the right sequence for your specific product depends on details a general guide cannot fully capture.

What transfers between dossiers

Whichever sequence you choose, the underlying evidence you build, verification and validation testing, usability engineering, elements of your quality management system, clinical or performance data, is not market-specific even though the submission logic and classification frameworks are. What a CE File Transfers to SFDA and the GCC covers this in more depth, including the additional MENA market consideration once you are planning beyond EU, UK, and US.

Status line and next review

This page carries a dated status line because its central insight, the UK legacy Class I arbitrage, is attached to a live regulatory reform process rather than a settled rule. Status as of 21 July 2026: the MHRA's draft Medical Devices (Amendment) Regulations 2026 has been published and does not, as currently drafted, reclassify SaMD from Class I to Class IIa; its confirmed software-specific provisions are a new international reliance pathway (anticipated around 2028), a mandatory UDI requirement, and a PCCP mechanism for SaMD. Overall adoption of the Draft Amendment is anticipated around December 2026, in force around June 2027, subject to Parliamentary approval. All of these forward dates — the draft's publication milestones, the ~2028 reliance pathway, and the adoption and in-force estimates — are estimates drawn from a live reform process, not confirmed dates, and we re-verify them against MHRA sources at each review. We will revise this page immediately on any material development in the UK regulatory process, or by 1 October 2026 at the latest, whichever comes first.

Two EU-side watch-items bear on the same comparison. First, in December 2025 the European Commission published a proposal to simplify the MDR and IVDR (COM(2025) 1023, procedure 2025/0404(COD)), which would, among other things, move much stand-alone software toward Class I self-declaration — a change that would narrow the very EU-vs-UK gap this page describes. It is a proposal, not law — adoption is realistically ~2027 at the earliest — so plan against the current rules while factoring the possible change into longer-term strategy. Second, the EU column carries costs the headline figures above don't isolate: EUDAMED registration has been mandatory since 28 May 2026, and AI-driven devices face the EU AI Act's high-risk conformity layer — due 2 August 2028 for medical devices under the Digital Omnibus postponement (Regulation (EU) 2026/1744, published in the Official Journal 24 July 2026, in force from 27 July 2026), which supersedes the original 2 August 2027 date.

Frequently asked questions

Is the UK legacy arbitrage available for every software product? No. It depends on the specific classification history and current transitional guidance applicable to your device, which changes over time. It is not a general rule that all Class IIa-equivalent software can self-certify in the UK; it applies to a specific set of legacy circumstances that need checking against your actual product, and its future is genuinely open — verify current status before relying on it.

Does launching in the UK first help my EU MDR submission? Indirectly. It does not substitute for MDR conformity assessment or shorten it, but real-world evidence and revenue generated from an earlier UK launch can strengthen the clinical evaluation and business case underpinning your subsequent EU submission.

Should we always launch in our biggest addressable market first? Not necessarily. This guide's core argument is that regulatory sequencing (which market's evidence and classification requirements are easiest to satisfy first) and commercial sequencing (which market has the most revenue potential) are separate questions, and the fastest path to your biggest market sometimes runs through a smaller one first.

How much does sequencing decisions actually save in practice? This varies enormously by product and is not something we can quote generically; the figures in this guide's comparison table are industry-cited ranges, not Venitara commitments, and your specific savings depend on your evidence maturity, device class, and which arbitrage opportunities currently apply to your product.

Is MENA a fourth market we should be sequencing alongside these three? Often, yes, particularly for teams with GCC investor or customer relationships. Saudi Arabia's SFDA approval is commonly understood in the region to function as a reference point for other GCC regulators, though we haven't found a single citable document formalizing this, so treat it as a plausible regional pattern rather than a guaranteed shortcut when you're sequencing. SFDA Medical Device Registration (MDMA) covers how Saudi Arabia's framework interacts with a CE-first strategy.

Does the UK's international reliance pathway change any of this sequencing advice today? Not yet. The pathway is anticipated around 2028, well after the rest of the Draft Amendment's expected entry into force, and it isn't in effect now. It's worth tracking if your roadmap extends that far out, particularly if you already hold or expect to hold Australian, Canadian, or US authorization, but it shouldn't change a sequencing decision you're making in 2026.

What happens to the UK arbitrage if the Draft Amendment doesn't get Parliamentary approval on the anticipated timeline? Then the current legacy framework, including the Class I self-certification pattern this guide describes, keeps applying until whatever does eventually change it. Delay in the reform process extends the arbitrage's runway rather than shortening it; the risk this guide flags is that reform happens and changes something, not that reform is imminent on a fixed date.

A structured conversation about your specific sequencing options, given your actual class, evidence base, and market priorities, is the right next step once you have a working sense of the landscape from this guide. This is not a decision a free tool can make for you responsibly, given how much rides on the current state of transitional rules and your specific evidence position.

---

Where next: MDR Annex VIII: How Device Classes Are Set · CE Marking Cost and Timeline for Medical Software · What a CE File Transfers to SFDA and the GCC · 510(k) vs De Novo vs PMA: Which FDA Pathway

Bring your device class and evidence base to one conversation, and we'll map the fastest honest sequence from there. Book an expert conversation →