Most applications succeed — but almost all get stopped along the way. A data-driven look at outcomes, delays, and what they mean for applicants.
The European Food Safety Authority has published more than 240 scientific opinions on novel food applications. We extracted every one we could obtain — 240 of the 241 issued — and ran a structured analysis across 239 of them. The picture that emerges is more nuanced than the headlines suggest.
Across the 239 opinions we analysed, 82% reached a favourable outcome — EFSA concluded the novel food was safe under the proposed conditions of use, most of them (119 of 239) with specific conditions of use attached. 13% were unfavourable, meaning the panel could not establish safety on the data provided, and a further 5% were inconclusive. Taken together, 18% ended non-favourably.
These numbers are broadly consistent with other analyses. Neytinck et al. (2025), who examined 107 post-2018 dossiers, reported an 85% favourable rate — the small difference is largely explained by our dataset including pre-2018 opinions assessed under the earlier regulation (Regulation (EC) No 258/97).
A favourable outcome does not mean a smooth process. In an analysis of post-2018 dossiers with available data, 91% received at least one additional data request from EFSA (Neytinck et al., 2025). The average was roughly 2–3 ADRs per dossier.
Each additional data request triggers a clock stop — a formal pause in the assessment timeline while the applicant prepares a response. Based on the available data, roughly half of the total time between submission and published opinion is spent in these clock-stop periods. That is time spent not waiting for EFSA, but waiting for the applicant to produce missing information.
The median total time from application to published opinion is around 16 months. The nine-month statutory deadline set by Regulation (EU) 2015/2283 has, in the available data, never been met.
The most commonly flagged areas in additional data requests are product characterisation and identity, production process documentation, and compositional data (Ververis et al., 2020). These are not obscure technical shortcomings — they are the foundational sections that establish what the product is, how it is made, and what it contains.
Toxicological data is another frequent trigger, particularly where the genotoxicity battery or the 90-day subchronic study is incomplete or where the test material does not match the product intended for the market.
The data tells a clear story: most novel foods get authorised, but the path is almost never straight. The question is not whether you will face an ADR, but whether you can minimise their number and respond efficiently when they come.
The most effective way to reduce clock-stop risk is to address the common gaps before submission. The patterns are well-documented — from EFSA's own published opinions, working group minutes, and guidance documents. Knowing what the panel looks for is not a guarantee against ADRs, but it shifts the odds meaningfully.
16% of applications are rejected outright as invalid before EFSA even begins its assessment (Neytinck et al., 2025). These are dossiers that fail basic completeness checks. That is an avoidable outcome.
- Neytinck, J. et al. (2025). Analysis of novel food application timelines and outcomes. *npj Science of Food*. - Ververis, E. et al. (2020). Novel foods in the European Union: Scientific requirements and challenges of the risk assessment process by the European Food Safety Authority. *Food Research International*. - EFSA (2024). Scientific Guidance for the preparation of applications for authorisation of novel foods. *EFSA Journal*. - Own analysis of 240 published EFSA opinions extracted via Europe PMC and Wiley; 239 structured (2003–2026).
Paste a section and Borgh will flag the gaps that matter — with references to the relevant EFSA guidance.
Try a section →