Estragole in Fennel Oil: What the Regulators Actually Said

Two sentences travel around the internet: "fennel oil contains a carcinogen" and "it's natural, so it's fine". The regulators said neither. They said estragole causes liver tumours in rodents at high doses, that the human relevance is plausible but unmeasured, and that exposure should be kept as low as practically achievable. This page walks through what was written, by whom, and when.

Fennel Essential Oil bottle

Fennel Essential Oil

Our own Fennel Essential Oil, in sizes from 0.5 oz to bulk.

From $13.99

Shop Now →

It is one of the supporting pages for our overview, is fennel essential oil safe, and it is deliberately about documents rather than fear.

Where estragole sits in sweet fennel oil

Estragole is a minor component of fennel oil, far behind trans-anethole. The European Medicines Agency's 2023 estragole statement tabulates it at 1.5% to 5.0% of sweet fennel oil and 3.5% to 12.0% of bitter, while noting the table is compiled mainly from food-sector sources, partly from older publications whose original data are no longer fully public (EMA estragole statement, 2023). A better anchor is the 2023 feed-additive opinion of the European Food Safety Authority, which analysed five batches of sweet fennel oil and found a mean of 3.76% estragole (range 3.60% to 4.10%) by one method and 4.42% by another, against a specification of 2.0% to 6.0% (EFSA FEEDAP Panel, 2023). The European Pharmacopoeia caps bitter fennel oil at 6.0%, as the EMA's 2024 oil statement records.

Those are other people's batches. Ours is analysed per lot by an independent laboratory, which quantifies both anethole and estragole, and we will send you both figures for your lot through the contact page. That is a process fact, not a claim that the number is low. How to read such a report is in reading a fennel oil label and GC-MS report, and why the sweet and bitter figures differ is in sweet vs bitter fennel oil. The oil's other aroma molecules are covered in the linalool connection.

What was actually measured

The key animal work came from the Miller laboratory in the 1970s and 1980s, in which estragole or its metabolites, given by diet, tube or injection, produced liver tumours in several mouse strains. The EMA describes the rat evidence as "suggestive, but indirect." A 3-month National Toxicology Program study found two of ten male rats at 600 mg/kg with multiple bile-duct cancers and a third with a liver adenoma, which the NTP authors read as significant evidence when taken with its other studies on related compounds. The EMA declined to use that study for limit-setting because it is not a properly designed cancer study. No human study exists (EMA estragole statement, 2023).

The mechanism is what carries the concern over to people. Estragole is converted in the liver to a reactive metabolite that binds DNA. That pathway has been shown in rodent and human liver cells in a dish, and the first step has been seen in people given estragole. The EMA reads that as making the rodent tumours "probably relevant" for humans. Two things it did not claim: that fennel oil itself has been tested for cancer (it has not, per the committee's 2024 public statement on the oil), and that anyone has measured cancer in people using fennel preparations.

What the regulators concluded, in order

The 2001 and 2002 positions disagree, and the EMA text reports both without ruling between them. All five statements are summarised in the 2023 EMA text, which is the source for this list. Note the shift in the EMA's own emphasis between 2005 and 2023: from no significant cancer risk for short adult use at recommended doses, to exposure "as low as practically achievable" with the admission that an exact limit cannot be defined.

The numbers regulators put on it, and what they are

The EMA derived a guidance value using the pharmaceutical-impurity method in the ICH M7 guideline: the rodent potency figure of 51.8 mg/kg per day, divided by 50,000 and multiplied by a 50 kg body weight, gives 0.052 mg per person per day, rounded to 0.05 mg. That is the value for adults, and also for pregnant and breastfeeding women. For children up to 11 years it is 1.0 microgram per kg of body weight per day. For comparison, the Council of Europe's 2005 estimate for estragole from all food is about 1 mg per person per day, and older estimates ran from 0.5 to 5 mg.

The committee is blunt about what these values are. It says that "an exact limit cannot be defined," that a conclusive assessment "seems not to be possible at current time," and that the guidance value is "a helpful tool" for judging sensitive groups, duration and dose. It also recommends a maximum of 14 days for medicines and says the same limits apply to skin products, restricted to intact skin and two weeks, "unless lower absorption rates" are shown. For skin, it says outright that no absorption data exist.

What it did to oral fennel oil

This is where the statement had teeth. In 2007 the committee had published a monograph for bitter fennel fruit oil as a traditional cough product at 200 microlitres a day. On 29 May 2024 it concluded that estragole intake at that dose made a monograph unsupportable, noting that other safer options exist, and proposed withdrawing it (EMA public statement, 2024). Back-calculating from its figures, that dose is about four drops, carrying roughly 11.5 mg of estragole at the 6.0% pharmacopoeia ceiling. Fennel tea is a different matter: a study cited in the statement found under 0.1% to 2.5% of the estragole reached an infusion, so the tea and the oil cannot be compared by weight of fennel.

The arithmetic for a bottle, labelled as ours

QuantityOilEstragole at 3.76%
One drop, undilutedabout 48 mgabout 1.8 mg
1 mL of a 2% blendabout 19 mgabout 0.72 mg
1 mL of a 1% blendabout 10 mgabout 0.36 mg
1 mL of a 0.5% blendabout 5 mgabout 0.18 mg
EMA guidance value, adult, per day0.05 mg

These assume a 0.05 mL drop, an oil density of 0.96 g/mL (implied by the EMA's own dose calculation) and EFSA's mean for someone else's batches. They are arithmetic, not a result. The unknown that matters is how much of the estragole on the skin reaches the bloodstream, which nobody has measured, so the table cannot say any blend is safe or unsafe. What it shows is why regulators spoke of "as low as practically achievable" rather than naming a number, and why we publish a dilution range with a patch test and do not pretend the guidance value is a threshold.

Where this leaves a reader

Nothing here is a finding that a diluted fennel oil product has caused cancer in a person. Nobody has tested that. What the documents support is a precaution that scales with how much and how long. For pregnancy, breastfeeding and children the EMA's values are the strictest and the sensible position is to skip the oil, which is also our published guidance. For regular long-term use, ask a pharmacist or physician and bring your lot figures. For occasional diluted use by a healthy adult, nobody can tell you the amount absorbed, and the documents describe a precaution, not a measured harm.

If you are worried, a toxicologist is the right reviewer for the science, and a pharmacist or physician is the right person for your own case. The practical overview, with who should skip the oil, is is fennel essential oil safe, and the medication and condition list is contraindications and interactions.

Fennel Essential Oil bottle

Fennel Essential Oil

Fennel Essential Oil from our own shop, in sizes from 0.5 oz to bulk.

From $13.99

Shop Now →

Sources & Further Reading