Remilk makes the clearest case study available, because it filed nearly everywhere and got four different answers.
The US Food and Drug Administration issued its no-questions letter in February 2023, roughly eleven months after filing, under GRAS Notice 1056. The Israeli Ministry of Health approved the same protein on 1 May 2023 — the first precision-fermentation dairy approval in that country. Singapore’s Food Agency cleared it too. The European Union application was filed, invalidated, resubmitted, and then withdrawn by the company.
Same molecule. Same safety dossier, broadly. Four regulators, four outcomes, and a spread of several years between the fastest and the one that never happened.
For anyone deciding where to file first, that spread is the whole decision.
The comparison
| Jurisdiction | Mechanism | Realistic timeline | Fee | Approvals to date |
|---|---|---|---|---|
| United States | GRAS notification to FDA, or self-affirmed GRAS | ~11–12 months to a no-questions letter; market entry possible immediately under self-affirmation | No FDA fee | At least seven fermentation-derived proteins since 2018 |
| Singapore | SFA pre-market approval | 9–12 months, stated by SFA, assuming a complete dossier with no clarification rounds | $1,750 per application | 14 novel foods on the first published list (17 March 2026), including precision-fermentation HMOs |
| Israel | Ministry of Health approval | Not published as a standard figure | Not published | Remilk approved 1 May 2023 |
| United Kingdom | FSA/FSS regulated products authorisation | ~2.5 years, with reforms underway to cut this | Varies | Precision fermentation identified by FSA as closest to market among emerging technologies |
| European Union | Novel Food Regulation (EU) 2015/2283, EFSA assessment | 2.5–4 years; nominal target is 18 months | Varies | Zero fermentation-derived animal proteins |
Read the timeline column as the actual planning input. The nominal figures published by regulators and the observed figures diverge sharply, and only in one direction.
United States — still the default first filing
The GRAS pathway remains the fastest route to revenue, for a structural reason: it does not require the regulator to say yes before you sell.
A company assembles a safety dossier, convenes an expert panel, and may self-affirm GRAS status and go to market while an FDA notification is pending — or without filing one at all. Remilk did exactly this, self-affirming in 2022 and selling in the US before its February 2023 no-questions letter arrived.
This is an honour system with regulatory backstop rather than a pre-market approval regime. Reasonable people disagree about whether that is good policy. What is not in dispute is that it compresses time-to-revenue from years to months.
The practical consequence: US clearance is the cheapest way to convert a laboratory result into a commercial reference. It also produces the safety dossier and the operating history that every subsequent jurisdiction will ask about.
Singapore — the underrated second filing
Singapore is frequently mentioned in passing as a fast alternative and rarely examined. The detail is more favourable than the reputation suggests.
The SFA states a review of approximately 9–12 months, on the condition that the dossier is complete and requires no further questions. That conditional matters — it is the same condition that stretches EU reviews into years — but the stated target is roughly half the EU’s, and the fee is trivial at $1,750 per application.
In March 2026 the SFA published its first List of Approved Novel Foods, covering 14 products approved between May 2019 and October 2025, including precision-fermentation human milk oligosaccharides. That list is more useful than it looks: it converts a previously opaque process into a public reference showing what has actually cleared, which is precisely the information a first-time applicant needs to calibrate a dossier.
Singapore also requires self-assessment checklists specifically for precision and biomass fermentation-derived products — a structured route to understanding what the regulator wants before submitting, which is exactly the pre-submission clarity the EU system lacks.
The practical consequence: for a company already holding US clearance, Singapore is a low-cost, moderately fast second market with a genuine consumer base and a regulator that publishes its outcomes.
Israel — fast, but poorly documented
Israel approved Remilk’s β-lactoglobulin on 1 May 2023, three months after the FDA no-questions letter and following what the Ministry of Health described as an in-depth assessment.
Beyond that, the public record is thin. There is no published standard timeline, no published fee schedule, and no equivalent of Singapore’s approved-products list that we could locate. The absence of a documented process is not evidence of a slow one — the Remilk sequence suggests the opposite — but it does mean a company cannot plan against a stated figure.
The practical consequence: likely fast for a well-prepared applicant with existing clearances, but requires direct engagement with the Ministry rather than desk research. Treat the timeline as unknown rather than short.
United Kingdom — slow now, reforming
Great Britain diverged from the EU regime after Brexit but inherited its shape, and with it much of its slowness. The current authorisation process runs to roughly two and a half years.
A meaningful share of that is procedural rather than scientific. Under the existing system a Statutory Instrument must be laid before an approval takes effect, which the FSA has said adds up to six months to a process that is already long. The reforms agreed by the FSA replace this with a public register of regulated products, and separately remove the requirement for already-authorised products to be reauthorised after a set period — freeing assessor capacity for new applications rather than re-litigating settled ones.
Both changes attack administrative overhead rather than scientific rigour, which is the correct target. Neither has yet produced an observable reduction in elapsed time for a precision-fermentation applicant.
Worth noting: the FSA and Food Standards Scotland have identified precision fermentation as the emerging food technology closest to market release. A regulator that has publicly named your category as the most imminent is a materially different counterparty from one that has not.
The practical consequence: not a first-launch market today. Worth filing into if you have a UK manufacturing or commercial footprint, and worth watching over the next 18 months to see whether the reforms move the observed figure.
European Union — the one to plan around, not for
The EU has authorised zero fermentation-derived animal proteins. Not one animal-free dairy protein, egg protein or fermentation-derived collagen can be sold in the bloc.
The two applications ever filed for such proteins — Perfect Day and Remilk, both for β-lactoglobulin — never reached scientific assessment. Perfect Day’s dossier sat in validation for over a year before being declared not valid in April 2024. Remilk’s failed validation on a study pre-notification technicality, passed on resubmission, and was then withdrawn by the company.
We covered the mechanics of why in our analysis of the EU approval gap — the suitability check, the transparency requirements, and the clock-stops that consume roughly half of total evaluation time.
The practical consequence: file when your intellectual property position can survive public disclosure and you can absorb a multi-year process of uncertain outcome. Do not build a commercial plan around EU revenue inside three years.
The sequencing that follows
The observed industry pattern is not a preference. It is arithmetic.
- United States first. Fastest route to revenue and to an operating safety record.
- Singapore and Israel next. Low cost, moderate timelines, real consumers, and both have now approved precision-fermentation products so the path is demonstrated.
- United Kingdom opportunistically, particularly if reforms bite and you have a local footprint.
- European Union as a long-horizon filing, resourced separately and never load-bearing in a near-term plan.
The uncomfortable implication for European companies is that this sequence is unchanged by where you are incorporated. Verley is French and holds a US no-questions letter for functionalised dairy proteins; its home market cannot buy the product. Europe was the strongest single geography for alternative protein funding in the twelve months to July 2026 — 62% of disclosed capital — while remaining a market where none of these proteins can legally be sold.
What we could not establish
Some honest gaps in the above.
Israel has no published timeline or fee schedule that we could find. The Remilk approval gives one data point, not a process.
The Singapore fee is stated as $1,750 in SFA guidance. We have assumed but not confirmed that this is Singapore dollars.
UK reform timing is unclear. The reforms were agreed and the changes were described as being enacted the following year, but we found no published figure for elapsed authorisation time under the new register, which would be the only meaningful test.
Fee schedules for the UK and EU vary by product type and applicant, and we did not find a single comparable figure. Fees are in any case the smallest cost in these processes — the binding constraint is elapsed time and the cost of capital across it.
What to watch
Does Singapore’s published list keep growing? A regulator that publishes outcomes and continues approving is a compounding asset for applicants. If the March 2026 list gets regular updates, Singapore becomes materially more attractive as a second filing.
Does the UK register produce a measurably shorter approval? The reforms target real overhead. The test is a precision-fermentation product completing authorisation in substantially under two and a half years.
Does any β-lactoglobulin dossier pass EU validation? No fermentation-derived animal protein has cleared this stage. The first that does — and how long its subsequent assessment runs — is the only signal that the EU position has genuinely changed.