Almost every guide to international expansion says the same reassuring thing: get an FDA clearance or a CE mark, and a lot of other markets will lean on it. That is true. It is also two completely different mechanisms wearing one word, and the difference decides something more important than cost.
In some markets a foreign approval is a shortcut. You can enter without one; having one means less dossier, less review, less money.
In others it is a prerequisite. It shortens nothing. It decides whether you are allowed to file at all.
A shortcut you lack is a line item. A prerequisite you lack is a closed market. Most planning treats them identically, and the mistake only surfaces after the budget is approved.
Singapore: the shortcut, working exactly as advertised
Singapore's HSA runs an abridged evaluation route recognising five authorities: the US FDA, EU notified bodies, Health Canada, Australia's TGA and Japan's MHLW/PMDA. Hold a current authorisation from any one of them and you submit a reduced-scope technical file instead of a full evaluation.
Crucially, you can still enter Singapore without any of them โ you just do more work. That is what a shortcut means. It lowers the price of entry; it does not control the door.
Malaysia works the same way and recognises more authorities than anyone else we track: the same five plus Singapore and Thailand. With a reference approval, the Malaysian conformity assessment body performs a verification audit rather than a full technical review.
Saudi Arabia: the same five countries, doing the opposite job
This is where the distinction stops being academic.
The SFDA used to run a GHTF reference route that let you lean on a US, EU, Canadian, Japanese or Australian approval directly. It cancelled that route at the end of 2021. Since January 2022, MDMA2 is the only pathway, and technical files from those authorities are no longer accepted as a reference shortcut.
But prior authorisation in a GHTF market remains a prerequisite for certification.
So Saudi Arabia still lists the same five countries as Singapore does. They have simply changed jobs. In Singapore they save you work. In Saudi Arabia they decide whether you have a market. A manufacturer holding none of the five cannot enter; a manufacturer holding four gets no discount for the extras.
Israel is built the same way. AMAR requires evidence that a device already has approval in a recognised country before it will register it โ the US, EU, Canada and Australia principally, with the UK, New Zealand and Switzerland also recognised. There is no independent Israeli assessment route to fall back on. No recognised approval, no Israeli registration.
Why this changes your sequence, not just your budget
If reliance were only ever a discount, sequencing would be simple: enter the biggest markets first and collect the discounts later. Because some markets are gated, the order inverts for a company that holds nothing yet.
- Holding neither FDA nor CE? Israel and Saudi Arabia are not "expensive" โ they are unavailable. They cannot appear anywhere in your plan until a gate-opening approval exists. Getting one is not a market entry; it is the precondition for a whole class of them.
- Holding one already? The shortcut markets are where it converts to money fastest โ Singapore, Malaysia, Thailand, Mexico. And the gated markets have quietly become available, which is a fact worth noticing rather than rediscovering in two years.
- Choosing which approval to get first? Ask which one opens the most gates rather than which market is largest. That is a different question and frequently a different answer.
Three conditions that no recognition list captures
Even where reliance is a genuine shortcut, the list of accepted authorities is not the whole rule.
Thailand requires the reference approval to be at least one year old. The Thai FDA accepts the GHTF five, the WHO and Singapore's HSA โ but the Concise Evaluation Route is only open to devices authorised for at least a year. A clearance you obtained last month does not qualify. There is a seasoning period between earning leverage and being able to spend it, and no table of recognised agencies shows it.
Thailand also runs an expedited route for products registered in more than one reference country, so approvals compound there rather than being interchangeable โ an argument for stacking approvals before filing rather than filing on the first one.
Mexico requires the filings to match. Its equivalence agreements recognise the FDA, Health Canada and Japan, and its own guidance states plainly that a foreign authorisation alone is not sufficient: the Mexican filing must align with the foreign one on device configuration, indications for use and manufacturing sites. A clearance covering a slightly different configuration, or a plant not named in the original submission, does not carry across. That mismatch is discovered after the route has been budgeted as the cheap one.
And reliance moves backwards. Saudi Arabia is the proof. Reliance is a per-market fact with a date on it, not a property of your FDA clearance โ and a plan built on pre-2022 Saudi experience is simply wrong today.
How the two categories actually split
We verified the reliance position for 28 of the 32 markets we cover, against the authorities' own published positions. The split is almost even, which is not what the "get FDA or CE and the world opens up" framing would predict:
| Reliance is aโฆ | Markets | Which |
|---|---|---|
| Prerequisite โ you cannot file without one | 8 | UAE, Israel, India, Nigeria, New Zealand, Philippines, Saudi Arabia, Taiwan |
| Shortcut โ cheaper and faster with one | 9 | Australia, Great Britain, Hong Kong, Mexico, Malaysia, Singapore, Thailand, Vietnam, South Africa |
| Neither โ independent assessment | 11 | Argentina, Brazil, Canada, China, Colombia, Costa Rica, EU, Japan, Korea, Peru, US |
Eight markets will not let a manufacturer file at all without a foreign approval in hand. That is a quarter of the world we cover, and it is invisible if you think of reliance purely as a discount.
Which approval opens the most doors
Counting only the markets whose position we verified, and counting a recognition once whether it functions as a gate or a discount:
| Reference authority | Markets recognising it |
|---|---|
| United States (FDA) | 15 |
| European Union (CE) | 15 |
| Canada (Health Canada) | 14 |
| Australia (TGA) | 11 |
| Japan (MHLW/PMDA) | 11 |
| Great Britain (MHRA) | 4 |
| Singapore (HSA) | 4 |
Two things in that table are worth more than the ranking itself.
The US and the EU are tied, and Canada is within one of them. A Health Canada licence carries almost exactly the same onward leverage as an FDA clearance โ 14 markets against 15. That is not the received wisdom, and for a company weighing which approval to pursue first it changes the calculation, because a Canadian licence is not the same price as a US one.
Singapore is a reference authority in four markets โ Malaysia, Thailand, Hong Kong and Australia all accept an HSA registration as leverage. An ASEAN-focused manufacturer that treats Singapore purely as a small market is missing what a Singaporean registration is worth elsewhere in the region.
One curiosity with a real implication: Brazil is a recognised reference authority in South Africa โ the only market we found that treats ANVISA as a reference. Given Brazil is otherwise one of the most expensive independent filings in the world, that a Brazilian registration buys leverage anywhere is worth knowing.
Two markets where approvals compound
Most reliance frameworks are indifferent to how many approvals you hold โ one qualifies you, the rest are wasted. Two are not.
Hong Kong's expedited scheme requires two or more valid reference approvals, plus a clean safety record: no reported deaths or serious injury, and no active recalls or field safety corrective actions. Hong Kong also recognises the widest set we found โ eight authorities, including Mainland China, South Korea and Singapore alongside the usual five.
Thailand runs an expedited route for products registered in more than one reference country. So in both markets a second and third approval each add value, which is an argument for stacking approvals before filing rather than filing as soon as the first lands.
What we checked, and what we got wrong
We are going to be unusually specific here, because we think it is the point.
Meridian's Regulatory Atlas held a reference-market list for 21 markets. When we set out to publish a league table of them, we checked first โ and found that 18 of the 21 had never been verified, that seven markets carried an identical five-country set, and that one market's list had been lifted from that country's pharmaceutical framework rather than its device framework.
So we verified them against the authorities' own positions before publishing. Nine of the markets we checked were wrong in some way. The worst two were ours understating the world, not overstating it:
- Australia we had recorded as recognising the EU alone. The TGA accepts market authorisation evidence from the EU, US FDA, Health Canada, Japan's MHLW/PMDA and Singapore's HSA โ and added UK certifications in 2026. Six, not one.
- South Africa we had recorded as the EU alone. SAHPRA's reliance pathways cover Australia, the US, the EU, Brazil, Canada, Japan and WHO prequalification.
- Taiwan we had listed as a five-market reliance regime. It is not a reliance market at all โ TFDA requires approval in the device's own country of origin plus a Free Sale Certificate, and other markets' approvals do not substitute.
- Malaysia we had listed China among the recognised authorities. It was never right; the extras beyond the GHTF five are Singapore and Thailand.
Four markets โ Kenya, Egypt, Ethiopia and Botswana โ claimed a reference route we could not verify to this standard. They are not in the tables above. We would rather publish 28 markets we have checked than 32 we have not.
The short version
- Establish whether each target market treats your approval as a shortcut or a prerequisite. They are opposite problems.
- If you hold no reference approval, the gated markets are not expensive โ they are unavailable. Sequence accordingly.
- Check the conditions, not just the list: age of approval, configuration match, manufacturing sites, which specific programme the reliance runs through.
- Re-check before you file. Saudi Arabia removed a route inside a single year, and nothing announced it to the companies who had planned around it.
Method, and what this is not
The market positions above were verified against the authorities' own published positions and regulatory reporting on 26 August 2026, and are held in Meridian's Regulatory Atlas alongside 30 other markets. Each carries the date it was checked and whether it has been through expert review.
Regulatory reliance frameworks change โ Saudi Arabia's changed inside twelve months, and Colombia is currently consulting on a new device regime that would add a reliance mechanism its own regulator has said explicitly is not fast-track. Verify against the authority before making a filing decision, and treat any reliance claim without a date on it as unreliable, including ours.
This is not regulatory advice.
See which markets your approvals actually open
Search 30 markets at once: where you're registered, where your competitors are, and which markets treat what you already hold as leverage. Free to start, no card required.
Search free โ