Is the erucic acid in British rapeseed oil a problem?

Is the erucic acid in British rapeseed oil a problem?

There is a bottle of cold-pressed British rapeseed oil in the kitchen, the golden kind from a farm shop or the supermarket's own, and it costs a fraction of what the cosmetic suppliers want for almond. Every spring you watch whole counties of the stuff go yellow from the train window. It is a mild, oleic-rich oil that behaves itself. The obvious thought follows: why am I not putting this in a balm?

And then you look it up, and the internet says rapeseed contains erucic acid, that erucic acid is restricted, that varieties were specially bred to get rid of it, and that there is a limit of 2 per cent — or 20 grams per kilogram, depending on which page you land on. Two different numbers in two different units for the same thing, neither of them explained.

Here is what that is actually about, in the order that matters. The limits you have been reading were written about eating. The thing that should worry a formulator is not the limit at all — it is that two genuinely different oils are sold under one name, and only one line on a specification tells you which one you have bought.

One name, two different oils

Rapeseed is the awkward corner of our guide to plant oils by fatty-acid group, because the botanical name settles almost nothing about composition.

The clearest place to see the split is the Codex Alimentarius Standard for Named Vegetable Oils — Codex being the joint FAO/WHO food standards programme, whose oil standard is the reference sheet almost every supplier's specification is written against. In the 2015 amendment of CXS 210-1999, rapeseed appears twice, as two separate products:

• "Rapeseed oil" (turnip rape oil; colza oil; ravison oil; sarson oil; toria oil), produced from seeds of Brassica napus L., Brassica rapa L., Brassica juncea L. and Brassica tournefortii Gouan.
• "Rapeseed oil – low erucic acid" (low erucic acid turnip rape oil; low erucic acid colza oil; canola oil), produced from low-erucic-acid seeds of varieties derived from Brassica napus L., Brassica rapa L. and Brassica juncea L.

Nearly the same plants. Two different product definitions, two different composition tables, one standard. And section 3.1 of that standard states the rule in a single line: "Low-erucic acid rapeseed oil must not contain more than 2% erucic acid (as % of total fatty acids)."

Erucic acid, if the name is new to you, is written C22:1 in the shorthand of fatty-acid tables: a twenty-two-carbon chain with one double bond. It is a very-long-chain fatty acid — long by the standards of an oil phase, where most of the chains are C16 and C18 — and it is monounsaturated, like oleic acid, just longer.

Here is what Table 1 of the standard allows each of the three relevant oils, as a percentage of total fatty acids. "ND" is the standard's non-detectable, defined there as 0.05 per cent or less.

Two things in that table matter more than the rest.

The erucic ranges are written so as not to overlap: above 2.0 and up to 60.0 per cent for the first entry, non-detectable to 2.0 per cent for the second. That is not an accident of sampling; it is the definition doing its job.

The oleic ranges, by contrast, overlap heavily — 8.0–60.0 against 51.0–70.0. So an oil declaring 55 per cent oleic could be either. Nothing else on the sheet separates them. Only the C22:1 line does.

At the bench: when you buy "rapeseed oil", ask which of the two Codex entries it is sold against, and get that in writing. A specification that names neither has not told you what is in the drum, however impressive the rest of the numbers look.

The number that frightened you was written about dinner

Now the part that clears up most of the confusion, and it is a question of which body of law you are standing in.

In Great Britain the restriction on erucic acid lives in food law. It sits in assimilated Regulation (EC) No 1881/2006 on contaminants in foodstuffs, in the Annex, Section 8 "Inherent plant toxins", entry 8.1 — "Erucic acid, including erucic acid bound in fat" — with three maximum levels:

• 20.0 g/kg for vegetable oils and fats placed on the market for the final consumer or for use as an ingredient in food, except camelina oil, mustard oil and borage oil;
• 50.0 g/kg for camelina oil, mustard oil and borage oil;
• 35.0 g/kg for mustard as a condiment.

Notice the units, because this is where the two numbers on the internet come from. Codex expresses erucic acid as a percentage of total fatty acids; the contaminants regulation expresses it as grams per kilogram of product. For a nearly pure triglyceride oil (a triglyceride being the fat molecule itself — three fatty acids on a glycerol backbone) the two are close in practice, but they are not the same measurement, and only the second has legal force, and only over a food sale.

Underneath those levels sits a dietary assessment, not a skin one. Russo and colleagues in Food Science & Nutrition (2021) record that the European Food Safety Authority set a tolerable daily intake for erucic acid of 7 mg per kg of body weight in 2016, and that the maximum permitted levels in the European Union were reduced on that basis from 50 to 20 g/kg for vegetable oils, and from 10 to 4 g/kg for infant and follow-on formula. Sissener and colleagues in Nutrients (2018) work from the same 2016 figure and report that low-erucic rapeseed oil and fish oil are the main sources in farmed-fish feed — and that a 200 g meal of mackerel at the highest concentration they measured would contribute 143 per cent of that daily intake for a 60 kg person. Kersting and colleagues (2022) modelled fourteen infant feeding scenarios against the same intake and found only three definitively below it.

Every one of those papers is about swallowing. None is about skin, and we make no claim in either direction about what topical use would mean.

So what governs erucic acid in a cosmetic in Great Britain? We looked rather than remembered. The full-text search facility at legislation.gov.uk returns 67 documents containing the phrase "erucic acid". We read all four pages of results. They are contaminants regulations, food and feed control regulations, infant formula regulations, food-contact materials regulations, customs tariff instruments, the 1976 Council Directive on maximum erucic acid levels in oils and fats, and — interestingly — two cosmetic ingredient-name instruments: the glossary of common ingredient names used for labelling, and the older inventory decision. The UK Cosmetics Regulation itself, EUR 2009/1223, does not appear in the list. The search reaches into annex material, too: the contaminants regulation carries erucic acid in its Annex, and it duly comes back in the results.

That is a negative finding and it should be read as exactly that, no wider. It does not say erucic acid is safe on skin, and it does not hand you a permission. What it does settle is the argument you will meet online, where a food contaminant limit in grams per kilogram is quoted at a formulator as though it governed a balm. It does not. What does govern your balm is what governs every cosmetic product placed on the GB market: its safety must be checked by a qualified safety assessor before you make it available, and the assessor decides what a given raw material can do in a given product at a given level.

At the bench: keep the two files apart in your head. A food figure tells you about the crop and the grade; your safety assessment tells you about your product. Quoting the first at the second is the commonest mistake in this whole subject.

Why there are two rapeseeds at all: the 1970s happened

The reason one crop carries two standards is a deliberate piece of plant breeding, and it is recent enough that a great deal of older writing about "rapeseed oil" describes a material most buyers will now never touch.

Azhar, Cahill and Khan review the history in Physiologia Plantarum (2025). Rapeseed was historically characterised by high erucic acid and high glucosinolate content, which restricted it mainly to industrial use; classical breeding in the 1970s produced the "double-low" canola varieties, cutting both, and turned it into a food-grade oil. Their account then runs on through hybrid cultivars, mutation breeding and marker-assisted selection to specialty oils with tailored fatty-acid profiles — high-laurate oils for industry, omega-3 enriched oils for nutrition — and on to genome editing with CRISPR/Cas9 for oil composition and pod-shatter resistance.

Two consequences follow for anyone reading around this subject.

An old table is not wrong, it is about a different oil. When a book gives rapeseed oil at 40-odd per cent erucic acid, it is describing the pre-1970s crop faithfully. Codex still keeps a standard for exactly that material, which is why the high-erucic entry has never been deleted: the crop still exists and still has industrial buyers. That is also why "rapeseed oil" from a supplier stays genuinely ambiguous until you have read the specification.

Within the low-erucic product, the actual figure varies a lot. Russo and colleagues bought 300 representative rapeseed oil samples in retail stores across Germany and measured them by gas chromatography with flame ionisation detection — the standard laboratory method for separating and quantifying the fatty acids in an oil. The results ran from 0.17 to 9.68 g/kg, with 241 of the 300 below 4 g/kg. Every sample was under the maximum level then in force and under the newer 20 g/kg one. But the highest result was about fifty-seven times the lowest, and all of them were equally entitled to say "compliant".

At the bench: "conforms to Codex" is a pass mark, not a number. If your formula depends on the figure — because you are documenting composition, or comparing lots, or answering a customer — ask for the lot value from the certificate of analysis, which is the per-batch test report a raw-material supplier can issue, not the generic specification sheet.

The British field, and what buying it from a supermarket costs you

Rapeseed is not an exotic here; it is one of the crops the country is made of. Defra's release on cereal and oilseed areas in England, updated 27 August 2026, puts the oilseed rape area at 276 thousand hectares, up 35.5 per cent — the first increase since 2023, though still below the 342 thousand hectares of that year. Which is why the yellow fields, and why British cold-pressed rapeseed oil sits on supermarket shelves in bottles that cost less than a fifth of what a cosmetic supplier charges for the same volume of something exotic.

That price gap is real, and so is the catch. A food bottle from a supermarket is made, tested and labelled as food: it comes with a best-before date and, ordinarily, nothing else. No certificate of analysis, no lot fatty-acid profile, no statement of which Codex entry it was produced against, no declared filtration or refining route, and no supplier who expects to be asked. For a batch you use yourself that may be a trade you are happy to make. For a product you intend to sell in Britain, the missing paperwork is the same paperwork your safety assessor will ask for, and "I bought it in Sainsbury's" is not a specification.

One more thing worth knowing before you decide it is not worth the bother: modern low-erucic rapeseed is an oleic-dominant oil, 51.0 to 70.0 per cent C18:1 by the Codex range. Whatever its name suggests, it sits sensorially in the same company as the group covered in oleic oils: olive, avocado, argan and hazelnut — which is also the honest answer to "is rapeseed oil good for skin". It is an ordinary, well-behaved, mid-weight oleic oil. The interesting question was never whether it is good. It is which of the two rapeseeds you have.

At the bench: if you want to try it, buy a cosmetic-grade lot with a certificate for the first side-by-side test, and compare it against an oleic oil you already know well rather than against an exotic. You are testing a workhorse, and it should be judged as one.

Mustard: also two products, also one name

Mustard is where very-long-chain fatty acids are not a historical hangover but the normal composition.

Codex defines mustardseed oil as derived from the seeds of white mustard (Sinapis alba L. or Brassica hirta Moench), brown and yellow mustard (Brassica juncea (L.) Czernajew and Cossen) and black mustard (Brassica nigra (L.) Koch), and gives it an erucic acid range of 22.0–50.0 per cent, with C20:1 at 5.0–13.0 and C24:1 at 0.5–2.5. Those are conformity ranges for authentic samples, not typical values.

The range is wide because the crop varies at cultivar level, and the size of that variation is startling. Sawicka, Kotiuk, Kiełtyka-Dadasiewicz and Krochmal-Marczak in the Journal of Oleo Science (2020) analysed oil from two cultivars of the same species, white mustard: 'Borowska' contained 22.2 per cent erucic acid, 'Bamberka' 3.8 per cent, with seed oil contents of 276 and 290 g/kg respectively. The authors conclude that the 'Borowska' oil suits technical purposes and the 'Bamberka' oil food purposes — on the strength of that one fatty acid alone. One species, one crop, a near six-fold difference in the number that defines the material.

And then the distinction that mustard writing most often gets wrong, which is not about erucic acid at all. Mustard "oil" means two chemically unrelated products.

A cold-pressed or expelled mustard seed oil is a fixed oil: triglycerides, of the composition tabulated above. A steam-distilled mustard essential oil is a volatile distillate with no fatty-acid profile whatsoever. Peng and colleagues in Food Chemistry (2014) isolated the essential oil from Sinapis alba seed by simultaneous steam distillation and extraction, identified fourteen components by gas chromatography–mass spectrometry, and found allyl isothiocyanate the main one at 71.06 per cent — the pungent compound behind hot mustard, and a very different thing to have in a product from a mild triglyceride oil. That finding is about their steam-distilled sample; mustard species differ in the glucosinolate chemistry that gives rise to it.

At the bench: if a listing says "mustard oil" without stating the method of manufacture, ask before you buy, and ask in those words — pressed or distilled. No fatty-acid profile will answer it for you, because the distillate does not have one, and the two materials need separate assessment.

What the long chains actually do to hand and skin

Less than you would like, on the published record, and we would rather label that than fill it.

We could not open a study that varied erucic acid content while holding everything else constant and measured how the oil felt. In our formulators' experience the two rapeseed materials do not behave interchangeably in an oil phase — a high-erucic oil can feel heavier and spread more slowly side by side, and mustard fixed oil, with a third of its chains at C20 and above, sits further along the same axis. That is bench observation, offered as a hypothesis for your own paired trial, not as a finding.

If you want the sensory question answered with numbers rather than adjectives, the measurements that do it are spreading value and polarity, and they live in why one oil feels light and another feels heavy — not in a fatty-acid table. A composition sheet is an identity document. It is not a prediction of feel.

Nor is it a prediction of shelf life. Erucic acid is monounsaturated; the fragile part of all three of these oils is alpha-linolenic acid, which Codex puts at 5.0–13.0 per cent for rapeseed, 5.0–14.0 for the low-erucic entry and 6.0–18.0 for mustardseed. On that count these oils keep company with the polyunsaturated group, and the measurements that matter are in oxidative stability of oils.

At the bench: treat a rapeseed or mustard lot as an oleic-to-polyunsaturated oil with an antioxidant and packaging problem, not as a curiosity with a long chain in it. The long chain is the identity question; the C18:3 is the storage question.

How to read the specification: six lines and one order

The whole subject reduces to a short reading protocol, every step of which is answered from documents you are entitled to ask for.

1. Which Codex entry is it sold against? "Rapeseed oil" and "rapeseed oil – low erucic acid" are two product definitions with two composition tables. Name neither and you have not described the material.

2. Read C22:1 first. Above 2 per cent of total fatty acids and it is not the low-erucic product, by the standard's own rule.

3. Corroborate with the neighbours. Codex caps the other long chains differently: C20:1 at 0.1–4.3 per cent for the low-erucic entry against 3.0–15.0 for the high-erucic one, C24:1 at ND–0.4 against ND–3.0. A sheet declaring C22:1 at 1 per cent alongside C20:1 at 8 per cent is internally inconsistent, and that is worth a query before purchase rather than a shrug.

4. Ask how it was measured, not only what it said. Peeters and Tamayo Tenorio in Foods (2022) compared extraction and derivatisation routes and found the Soxhlet procedure yielded the highest erucic acid concentrations — and, the finding with teeth, that with the Folch procedure some natural cis erucic acid converted to its trans isomer, brassidic acid, so that ignoring the conversion can lead to underestimation of erucic acid content. Their samples were rapeseed protein products rather than oils, and we flag that limit; the lesson about method transfers anyway.

5. Use the identity checks if the profile looks odd. The Appendix to CXS 210 gives supplementary authenticity factors: the Crismer value for low-erucic rapeseed oil should be in the range 67–70, and its brassicasterol content greater than 5 per cent of total sterols — sterols being the minor components an oil carries alongside its triglycerides, and brassicasterol being the one characteristic of Brassica species. Codex is explicit that these are supplementary: a product meeting the essential factors but not these may still conform. They are fingerprints, not pass marks.

6. Do not read one certificate as the crop. Assess the lot you were sent, from its own specification and certificate. How much harvests move is the subject of how climate changes oil composition.

At the bench: keep the certificate with the lot number in the same folder as your batch record. When something behaves unexpectedly six months later, the question "which rapeseed was that?" has a documented answer or it has none.

Where the limit actually sits

The limit is not where the internet puts it. There is no erucic acid entry to comply with in the GB cosmetics regulation — we searched for one. There is a food contaminant limit, in grams per kilogram, built on a dietary intake assessment, resting on a breeding programme from the 1970s that split one crop into two standards, one of which is still legally sold for industrial use.

A formulator's job here is not compliance with a food number. It is knowing which of the two oils is in the drum — and the only line on the certificate that answers that is C22:1.

Sources

• FAO/WHO Codex Alimentarius Commission. Standard for Named Vegetable Oils, CODEX STAN 210-1999. Adopted 1999; revised 2001, 2003, 2009; amended 2005, 2011, 2013 and 2015 (2015 amendment consulted). CXS_210e_2015 (PDF) — consulted 22 September 2026.
• Commission Regulation (EC) No 1881/2006 setting maximum levels for certain contaminants in foodstuffs, as it has effect in UK law; Annex, Section 8, entry 8.1. legislation.gov.uk — consulted 22 September 2026.
• Full-text search of legislation.gov.uk for the phrase "erucic acid", 67 results over four pages, all inspected. legislation.gov.uk — searched 22 September 2026.
• Department for Environment, Food and Rural Affairs. Cereal and oilseed areas in England, updated 27 August 2026. gov.uk
• Office for Product Safety and Standards. Making cosmetic products available to consumers in Great Britain. gov.uk — consulted 22 September 2026.
• Azhar M., Cahill D. M., Khan G. A. A brief history of canola genetic gains: from classical breeding to genome editing. Physiologia Plantarum 177(6), e70644, 2025. PubMed 41255332
• Russo M., Yan F., Stier A., Klasen L., Honermeier B. Erucic acid concentration of rapeseed (Brassica napus L.) oils on the German food retail market. Food Science & Nutrition 9(7), 3664–3672, 2021. PubMed 34262725
• Sissener N. H. et al. Erucic acid (22:1n-9) in fish feed, farmed, and wild fish and seafood products. Nutrients 10(10), 1443, 2018. PubMed 30301170
• Kersting M., Kalhoff H., Honermeier B., Sinningen K., Lücke T. Erucic acid exposure during the first year of life — scenarios with precise food-based dietary guidelines. Food Science & Nutrition 10(1), 115–121, 2022. PubMed 35035914
• Sawicka B., Kotiuk E., Kiełtyka-Dadasiewicz A., Krochmal-Marczak B. Fatty acids composition of mustard oil from two cultivars and physico-chemical characteristics of the seeds. Journal of Oleo Science 69(3), 207–217, 2020. PubMed 32051354
• Peng C. et al. Chemical composition, antimicrobial property and microencapsulation of mustard (Sinapis alba) seed essential oil by complex coacervation. Food Chemistry 165, 560–568, 2014. PubMed 25038712
• Peeters K., Tamayo Tenorio A. Comparing analytical methods for erucic acid determination in rapeseed protein products. Foods 11(6), 815, 2022. PubMed 35327237