An FL number is written as two digits, a dot, and three digits. Vanillin is 05.018. Ethyl acetate is 09.001. Limonene is 01.001.
The leading pair sorts substances by chemistry, and it is useful once you can read it. It is also the source of a specific and expensive confusion, because the EU flavoring rules contain two chemical group schemes and they are not the same scheme. Matching an FL number prefix against the wrong one produces a confident answer to a question you did not ask.
Two schemes, different jobs
The one people usually mean sits in Annex I of Commission Regulation (EC) No 1565/2000, the regulation that set up the evaluation programme. Its heading is "Chemical groups for flavouring substances" and it lists 34 of them, each defined in a sentence or two of structural chemistry. Group 1 is straight-chain primary aliphatic alcohols, aldehydes, acids, acetals and esters, with a clause excluding aromatic or heteroaromatic moieties. Group 34 is amino acids. A footnote explains the purpose: these groups "are expected to show some metabolic and biological behaviours in common." They exist so that substances could be assessed together rather than one at a time.
The FL number prefix is a different thing. The 2,475 entries in Annex I Part A of Regulation (EC) No 1334/2008 use 17 distinct prefixes, 01 through 17. Seventeen is not thirty-four, so whatever the prefix is, it is not the 1565/2000 chemical group.
That matters because the two look interchangeable. Both are small integers attached to flavouring substances in EU law, both are called chemical groups in ordinary conversation, and a substance can plausibly carry a low number in each. Reading an evaluation written against one scheme while holding an FL number from the other is a category error that produces no error message.
What the seventeen prefixes actually hold
The regulation that publishes the Union list does not set out the prefix scheme in its text, so what follows is a description of the 2,475 entries as they stand rather than a quotation of an official definition. The pattern is consistent enough to be useful.
| Prefix | Entries | What the entries are |
|---|---|---|
| 01 | 35 | Hydrocarbons. Limonene, p-cymene, the pinenes |
| 02 | 206 | Alcohols. Propan-1-ol, butan-1-ol, isopentanol |
| 03 | 17 | Ethers. 1,8-Cineole, dibenzyl ether |
| 04 | 79 | Phenols and phenol derivatives. Eugenol, isoeugenol, guaiacol |
| 05 | 172 | Aldehydes. Acetaldehyde, propanal, vanillin |
| 06 | 114 | Acetals. 1,1-Diethoxyethane, citral diethyl acetal |
| 07 | 195 | Ketones. Heptan-2-one, acetophenone |
| 08 | 111 | Acids. Formic, acetic, propionic, lactic |
| 09 | 789 | Esters. Ethyl acetate and 788 others |
| 10 | 59 | Lactones. Nonano-1,4-lactone, the macrocyclic musks |
| 11 | 17 | Amines. Butylamine, propylamine |
| 12 | 225 | Sulfur compounds. Methanethiol, the methylthio aldehydes |
| 13 | 142 | Furans. 5-Methylfurfural, the furoate esters |
| 14 | 124 | Nitrogen heterocycles. Pyrazines, pyridines, pyrimidines, quinoxalines |
| 15 | 84 | Sulfur heterocycles. Thiophenes, thiazoles |
| 16 | 75 | Everything else |
| 17 | 31 | Amino acids. Alanine, arginine, aspartic acid |
The distribution says something about the trade on its own. Esters are 789 of 2,475, almost a third of the entire Union list, which is what you would expect of a list built around fruit and confectionery flavor. Sulfur compounds and their heterocycles together come to 309, which is the savory and roasted end of the business. Amines number seventeen.
Group 16 is where the rules are
Twenty-five entries on the Union list carry restrictions on use, meaning a limit expressed in milligrams per kilogram against specific food categories rather than a plain authorization.
Seventeen of those twenty-five are in group 16, the miscellaneous one.
That is not a coincidence, and it is the most useful thing the prefix tells you. Group 16 is where substances land when they do not fit a functional-group box: ammonium salts, flavonoids such as naringin and hesperetin dihydrochalcone, ferric ammonium citrate, and the modern cooling and taste-modifying compounds with names that run to a full line of type. Those are precisely the substances whose use the EU has wanted to bound. The tidy functional-group classes, the esters and aldehydes and ketones that make up the bulk of the list, are almost entirely unrestricted: 789 esters carry no restrictions between them.
So a formulator scanning FL numbers has a crude but real signal. A 16 is worth reading carefully. A 09 almost certainly is not, though the entry is still the thing that decides it.
Who actually did the evaluating
The 34-group scheme was built for a programme, and the programme left a mark on every entry in the Union list.
Article 2 of Regulation (EC) No 1565/2000 says a substance in the register need not be re-evaluated if it had already been classified by the SCF in Category 1, by the CEFS in Category A, or by JECFA as presenting no safety concern at current intakes, provided purity and specification information is supplied. Rather than start from zero on two and a half thousand substances, the EU accepted existing work and concentrated effort on what was left.
You can see the result in the entries themselves. Of the 2,475 on the list, 1,831 record their evaluation as EFSA's, 573 as JECFA's, and the remainder name the SCF or the Council of Europe alone or in combination. The Council of Europe appears on five entries by itself. Those attributions are not decoration: they tell you which body's published opinion is the one to read when you need the reasoning behind an authorization, and the small groups are the old ones, carried across rather than re-examined.
What the prefix does not tell you
It does not tell you the 1565/2000 chemical group, for the reason above.
It does not tell you whether a substance is currently authorized. The Union list has moved since Commission Implementing Regulation (EU) No 872/2012 adopted it, and 78 substances that appeared in the 2012 text have since been withdrawn. They keep their FL numbers in the historical record and are no longer on the list. A prefix is stable; membership is not.
It does not tell you a purity specification applies. 209 entries carry one and the rest do not, spread across every prefix.
And it does not imply anything about a substance absent from the list. The Union list covers flavouring substances. It does not cover flavouring preparations, thermal process flavourings, smoke flavourings or flavour precursors, each of which the regulation treats separately. A substance with no FL number at all may be perfectly lawful in the EU under another instrument.
Reading the list yourself
Two practical notes. Read the consolidated text of Regulation (EC) No 1334/2008 rather than the 2012 adopting regulation, which has been amended repeatedly and is misleading on its own. And when an entry carries a restriction, the restriction is written against the EU food category numbering, so "category 5.3" and "category 14.1" are references to a separate scheme you will need beside you.
You can search all 2,475 entries by FL number, CAS number or name in our free EU Union List lookup, which reports the 78 withdrawals as withdrawn rather than as absent, and shows the restrictions and purity specifications where the regulation gives them. The companion guide to FL numbers covers the two defaults hiding in the list's columns.