The EatMatrix library / 356 additive guides

Food additives, explained.

What they do in your food. What the evidence says about your health. Find an additive for a safety summary, research and country rules.

Search by E number or ingredient name

356 entries

Foods, vitamins & extracts →

About the ratings & search

Badges show EatMatrix’s reading, with the basis explained in each guide. “Safe in moderation” applies to the assessed food uses and amounts. “Use caution” flags a specific exposure, population or evidence limitation. Country permission and the safety of a whole product are separate questions.

Enter one code or ingredient name. An exact E number matches an entry's own code. E160b(i) and E160b(ii) open the grouped annatto entry; E460(ii) opens the grouped cellulose entry when no separate powdered-cellulose entry exists.

What do E numbers mean? · Our evidence standard · Study types explained

E100–E199 · Colours

E200–E299 · Preservatives

E300–E399 · Antioxidants & acidity regulators

E400–E499 · Thickeners, stabilisers, emulsifiers

E500–E599 · Acidity regulators & anti-caking agents

E900–E999 · Glazing agents, gases, sweeteners

How to read E numbers on a food label

What does an E number mean?

An E number identifies a food additive in the European Union. Additives serve functions such as preserving food or changing its colour or texture. The code identifies the substance; it does not describe how much is in a particular product.

Why does a label show a name instead of an E number?

EU labels identify the additive’s function and either its name or E number. For example, a label can list xanthan gum or E415. Search this catalogue by either code or ingredient name.

Are EatMatrix ratings the same as legal approval?

No. Each guide states whether its answer is a dated reading of selected authority assessments or a recorded database rating. Neither is permission for every product. A specific use needs to be checked against the rules for the substance, food and territory. Missing information is not evidence of a ban or a safety conclusion.

EU label explanation: EFSA’s food additives guide (reviewed 13 April 2026; checked 2 October 2026). The catalogue is not a complete list of current EU authorisations.

Ratings, research and country permissions: three different checks

A useful ingredient entry needs to explain why the additive is used in food, what research says about effects on people, and whether a specific use is permitted in a market. These questions need different sources. An ingredient name alone does not tell us the amount in a product or its effect on a particular person.

Recorded EatMatrix rating: calculated from imported database labels. It is not a verified current scientific review. A database update date does not establish when an assessment was reviewed.
Source-based safety reading: 89 previously unrated entries now have a dated reading of selected authority assessments, key exceptions and practical advice. Each identifies the sources and assessment years. This is not a systematic review of every study or a claim that all uses are safe. If an entry’s identity changes, it needs a new evidence review.
Research: use the correct ingredient, ingestion route, dose, population and outcome. Distinguish human trials, observational associations, animal experiments and cell experiments. A new publication date alone does not make a study more relevant or stronger.
Permission: check the exact substance, territory, food category, limits and effective dates. A scientific opinion or GRAS use is not interchangeable with an unconditional approval. A missing listing is not automatically a ban.

Where to check the market rules and assessments

These are source directories, not completed checks for every ingredient. The UK needs territory-specific checks; FAO/WHO JECFA provides international scientific evaluations rather than national permissions. Japan's CAA took over food-safety standards responsibilities from MHLW on April 1, 2024.

Markets, authorities and the purpose of each source
Market	Authority and source
European Union	European Commission / EFSA Use permissions / scientific assessment
United States	FDA Food-additive regulations and GRAS uses
England / Wales / Northern Ireland	FSA Code guide; check the applicable territory and conditions
Scotland / Great Britain rules	Food Standards Scotland Use the GB authorisation register and its linked conditions
Canada	Health Canada Permitted food additives and their conditions
Australia / New Zealand	FSANZ Food Standards Code use permissions
Japan	CAA / FSCJ Use standards / scientific assessment
International assessment	FAO/WHO JECFA Scientific evaluation, not national permission

Source-directory and catalogue coverage audit: October 4, 2026. Ingredient-specific current legal and scientific verification remains pending. The linked source's own dates and conditions matter.

What kind of evidence is this?

The label tells you how the evidence was collected. It is not a score. Confidence also depends on how the work was done, how precise the result is, and whether other studies agree.

Randomized trial: People are assigned to groups at random. Read the result for the ingredient, amount, participants and outcome actually tested. Randomisation alone does not make a study conclusive.
Human intervention: Researchers give people an ingredient or change their diet. Check whether there is a comparison group and whether assignment was random.
Observational study: Researchers track what people eat and what happens to their health. A link between the two does not by itself show cause and effect; other differences may help explain it.
Laboratory study, not an ingestion trial: Experiments in cells, animals or other laboratory systems can explore mechanisms. They do not directly establish what happens to people eating the ingredient.
Dietary exposure estimate: Estimates how much people may consume. An estimated intake is not a measured illness or proof that the ingredient caused harm.
Regulator review: An authority assesses evidence for specified uses. Read its conclusion, conditions and date; a scientific assessment and permission to use an additive are separate decisions.
Expert critique: An expert examines another study’s methods or interpretation. This can identify limitations, but it is not an independent repeat of the original experiment.
Preprint, not peer-reviewed: A report released before peer review. Its findings may change, and its study design still needs to be assessed.

Further reading: Cochrane’s guide to assessing confidence in evidence and NIH’s guide to understanding health research. These explanations are reading aids; EatMatrix has not completed a GRADE assessment for these entries.

Read the whole label.

Scan a food label with EatMatrix on iPhone to look up its ingredients in the app.

Get EatMatrix for iPhone