Every international medical device manufacturer eyeing the Australian market eventually asks some version of the same question: do we go straight to the TGA, or do we get FDA clearance or EU CE marking first and bring that evidence to the TGA under its reliance pathway?

It’s a reasonable question, and there’s no shortage of confident answers to it — usually along the lines of “the US is faster” or “just do Australia first, it’s quicker.” Both of those claims are true in a narrow sense and misleading in the way that matters most. The right sequence depends on what your device actually is, not on a general reputation one regulator has for being slow or fast. This is especially true for AI-enabled and software-based devices (SaMD), where classification and evidence requirements can diverge sharply from the “traditional” device playbook.

Below is a balanced look at the three broad options, the trade-offs each one carries, and — more importantly — the questions that should actually determine your sequencing decision, rather than a rule of thumb.

The Real Question Isn’t “Which Regulator Is Easier”

Speed and difficulty are not the same axis, and conflating them is where most informal strategy conversations go wrong.

TGA operates a Comparable Overseas Regulator (reliance) pathway, allowing a manufacturer to bring evidence from a small list of recognised regulators — including the US FDA and EU MDR Notified Bodies, among others — to support an abridged Australian application. On paper, this makes “get cleared somewhere internationally first, then bring it to the TGA” an attractive default. But which international route to take first is a genuinely open question, and the honest answer is: it depends on your device’s classification and evidence base, not on which market has a reputation for moving faster.

Three Paths, Three Trade-Off Profiles

Direct to TGA. This is the fastest way to get a first market authorisation in Australia, and for some companies that matters — investor milestones, an existing distribution relationship, or a need for early revenue. But going direct doesn’t reduce the work required for larger markets later, and Australia is not itself recognised as a comparable regulator anywhere else. If the US and EU are both on your roadmap, a TGA-first approval buys speed in one market without shortening the path in any other.

FDA first. The US 510(k) pathway is well trodden for many device categories, and the Medical Device Single Audit Program (MDSAP) lets a manufacturer complete a single QMS audit that’s accepted by TGA, Health Canada, Japan’s MHLW, and Brazil’s ANVISA — a genuinely reusable piece of evidence, not just an American formality. FDA has also built a comparatively mature AI/ML-specific framework (Predetermined Change Control Plans, Good Machine Learning Practice) for managing iterative model updates post-clearance. The catch, covered in more detail below, is that all of this speed assumes your device qualifies for the 510(k) route in the first place.

EU MDR (CE marking) first. The EU is not part of MDSAP, so a Notified Body audit of your quality system is a separate, non-transferable certification effort — one more reason EU-first tends to take longer end to end. Notified Body capacity has also been under real strain: the European Commission’s own 2026 survey work has flagged extended queues for higher-risk device classes. Under a recent regulatory reform (Commission Implementing Regulation (EU) 2026/977, adopted 4 May 2026), the EU has moved to impose binding maximum timelines on Notified Body conformity assessments — a welcome development, though it will take time to show up in practice. Separately, the EU AI Act’s high-risk obligations for AI embedded in medical devices were pushed back to 2 August 2028 following the “Digital Omnibus” reform package (finalised by the Council on 29 June 2026), giving manufacturers more runway on that specific overlay than the original 2027 deadline implied.

The Assumption That Trips Manufacturers Up

Here’s where a lot of pathway comparisons — including, frankly, a lot of the “US is just faster” shorthand — quietly go wrong: they assume a predicate device exists for the FDA 510(k) route. If it does, 510(k) review timelines of roughly three to twelve months are realistic, and the pathway lives up to its reputation. If it doesn’t, the fallback is FDA’s De Novo pathway, and the comparison changes materially.

De Novo has an official FDA target of 150 days, with real-world timelines often closer to 250 days once information requests are factored in, plus a meaningful user fee and a heavier evidence-generation burden than 510(k) — because there’s no substantial-equivalence shortcut to lean on. That still tends to be faster than a constrained EU Notified Body queue, but the gap narrows considerably, and “established, fast US pathway” stops being a safe assumption.

This matters most for genuinely novel AI functions or unusual combinations of claims (for example, an imaging-analysis function paired with a clinical decision-support function in a single integrated product). Radiology AI as a category is now well precedented at FDA — computer-aided detection/diagnosis clearances make up a large share of AI/ML medical device clearances — but a well-precedented category doesn’t guarantee a predicate for your specific, exact intended use. The only reliable way to know is to ask FDA directly, early, through a Pre-Submission (Q-Sub) meeting, rather than assume.

It also means the “EU is slower, so EU is harder” framing deserves a caveat. Under EU MDR’s Rule 11, software that informs diagnosis or treatment decisions is likely to land in Class IIb, and Class III isn’t out of the question where the underlying condition carries a risk of death or irreversible harm. At that classification, the EU requires full Notified Body review of your technical documentation and genuine clinical evaluation evidence — not a substantial-equivalence shortcut — plus ongoing post-market clinical follow-up and periodic safety reporting. If your FDA route turns out to require De Novo rather than 510(k), the evidentiary gap between the US and EU routes narrows on both timeline and rigour. The honest takeaway isn’t “EU is worse” or “US is easier” — it’s that both routes converge in difficulty once your device sits outside the comfortably precedented middle of the road, and the deciding factor is your specific classification and evidence position, not the regulator’s general reputation.

MDSAP: The One Lever That Works Regardless of Sequence

Whichever market you approach first, MDSAP certification is worth pursuing early and independently of that decision. It’s accepted as QMS evidence by TGA (via desktop review, without a separate on-site TGA audit), Health Canada, Japan, and Brazil — genuinely reusable across multiple reliance-style pathways. It is not accepted by the EU, and it is not itself a marketing authorisation in any market — a 510(k), De Novo grant, PMA, or CE mark is still required on top of it. But as a piece of infrastructure that de-risks and speeds up several markets at once, it’s close to a free strategic decision rather than a market-specific bet.

Questions to Answer Before You Commit

Rather than defaulting to a rule of thumb, work through these before locking in a sequence:

  • Does a predicate exist for your device’s specific intended use — not just its general category? This is worth confirming directly with FDA via a Pre-Submission meeting before you commit to a timeline.
  • How does your device classify under EU MDR Rule 11, and does that classification trigger full Notified Body technical file review?
  • Do any components of your product (clinical decision-support features, for instance) potentially qualify for narrower regulatory treatment or exclusions under any of the jurisdictions you’re targeting?
  • Do you have a genuine near-term commercial reason to prioritise speed to market in one jurisdiction — investor milestones, an existing distribution deal, working capital — that should outweigh a “build once, reuse everywhere” sequencing logic?
  • Which markets are actually on your three-to-five-year roadmap? A pathway choice that’s optimal for a US-and-Australia strategy may not be optimal if the EU is a near-term priority too.
  • Is your technical documentation being built in a modular, IMDRF-aligned format from the outset, so it’s reusable across submissions rather than rebuilt for each market?

The Bottom Line

There’s no universally correct sequence, and any advice that says otherwise is skipping the classification and predicate-availability work that should actually drive the decision. FDA’s pathway is often faster — when a predicate exists. The EU’s pathway is often slower — but its rigour doesn’t disappear just because the US route happens to be quicker for a given device. TGA-direct is fastest in Australia specifically, but doesn’t reduce the work needed anywhere else. The right call comes from mapping your specific device against each jurisdiction’s classification rules and evidence requirements before you commit — not from a generalised view about which regulator is “easier.”

This article reflects the regulatory landscape as understood in July 2026, including recent developments such as the EU’s Digital Omnibus AI Act amendments and Implementing Regulation (EU) 2026/977 on Notified Body timelines. Medical device regulation is an active, fast-moving area — this is general information, not legal or regulatory advice for any specific product. Manufacturers should seek a formal classification and pathway assessment for their specific device before committing to a regulatory strategy.