Horse Chestnut: Raw Seed Poisoning and Esculin
Every autumn, horse chestnut trees drop thousands of shiny brown seeds onto pavements, playgrounds and park paths. They look like the sweet chestnuts sold roasted on street corners. They are not the same thing, and they are not food. Poison-control services describe all parts of the horse chestnut tree as toxic, with the seeds the greatest concern simply because they are the part people pick up and eat.
This page covers what is known about raw horse chestnut poisoning: what the poison-centre case reports describe, how to tell the seed from an edible chestnut, the unusual exposures recorded in the medical literature, and the confusing question of which compound — esculin or aescin — is to blame. It also explains why none of this contradicts the safety record of the standardised seed extract tested in clinical trials. The two are different exposures.
Table of Contents
- What the Authorities Say
- What Eating Raw Seeds Does
- Two Poison-Centre Calls
- Horse Chestnut or Sweet Chestnut?
- Unusual Exposures in the Medical Literature
- Esculin or Aescin: Which Is the Poison?
- Esculin in the Laboratory
- Why the Extract Is a Different Story
- The Coumarin and Bleeding Question
- Key Facts at a Glance
- Key Research Papers
- Connections
What the Authorities Say
- US National Center for Complementary and Integrative Health (NCCIH): "The raw seeds, bark, flowers, and leaves of horse chestnut are unsafe when used orally because they contain a toxic component." NCCIH names that component as esculin, and states that standardised seed extracts from which it has been removed are likely safe for short-term use.
- US National Capital Poison Center: all parts of the tree are toxic, the greatest concern is the seeds, and eating them can cause serious stomach upset.
- European Forest Atlas (European Commission): unprocessed seeds are poisonous.
- European Medicines Agency: its herbal monograph covers only processed seed extracts, as standardised dry extracts for oral use and specified extracts in creams and gels. Raw seed is not among the listed preparations.
The sources agree on the practical point. Where they diverge is the chemistry, covered below.
What Eating Raw Seeds Does
The National Capital Poison Center's account, drawing on the published case literature, describes the following:
- Most common: stomach irritation — stomach pain, nausea, vomiting or diarrhoea.
- Timing: symptoms most often develop 15 minutes to 3 hours after eating, and can last 2 to 3 days.
- Children: reports of children who, in addition to major stomach irritation, felt weak, had difficulty walking, became drowsy, showed muscle twitching or experienced paralysis.
- Severe adult cases: a review of serious plant poisonings reported to the Swiss toxicology information centre over 29 years (1966–1994) included three horse chestnut cases with severe symptoms within 15–30 minutes of swallowing 1 to 4 seeds: redness, facial swelling, massive vomiting, low blood pressure and collapse. One patient developed angioedema — swelling of the lower face that blocked the airway.
The same article explains why most cases stay mild: the toxin is not absorbed very well.
Two Poison-Centre Calls
The poison center's article includes two anonymised calls that show what ordinary exposures usually look like:
- A mother and her 8-year-old son each tried a nut given to the boy, believing it edible. It tasted very bitter, they spat it out and called the poison center. Neither developed symptoms.
- A man boiled and ate four horse chestnuts, then realised what they were. He had an upset stomach that lasted about four hours.
Boiling did not make the seeds safe in the second case. The bitterness that stopped the first family is a natural warning: edible sweet chestnuts taste starchy, horse chestnuts are intensely bitter.
Horse Chestnut or Sweet Chestnut?
Mix-ups happen because the shiny brown seeds look alike. The poison center lists the differences:
- Husk. Edible (sweet) chestnuts grow in a dense, spiky husk that looks furry. Horse chestnut husks have fewer, more widely spaced spikes.
- Number of nuts. A sweet chestnut husk holds several nuts. A horse chestnut husk usually holds one.
- Shape. Sweet chestnuts have a pointed end or a small tassel. Horse chestnuts are completely round and leathery, with a large pale spot (the hilum, where the seed was attached).
- Taste. Sweet chestnuts are starchy. Horse chestnuts are very bitter.
The trees are not closely related: sweet chestnut is Castanea, while horse chestnut is Aesculus, with big hand-shaped leaves of five to seven leaflets. The North American buckeyes are close relatives of horse chestnut, and LiverTox notes that the California and Ohio buckeyes are different species from it; the poison center notes that the pale spot is why horse chestnuts themselves are sometimes called "buckeyes".
Unusual Exposures in the Medical Literature
The European Medicines Agency's 2020 assessment report collected several case reports involving horse chestnut preparations that were not standardised extracts. It judged them not relevant to its monograph on ethanol extracts, but they show what other forms of exposure can do:
- Snuff. A 15-year-old in Poland was admitted to hospital with vomiting, breathlessness, burning in the nose and throat, and fainting after sniffing powdered horse chestnut seeds as snuff, a local tradition in the Kashubian region. Laboratory tests were normal, and she went home after two days. The case authors concluded that nasal use of powdered seed can cause sudden but self-limiting symptoms.
- Unpurified paste. A 32-year-old man who had eaten three boxes of a traditionally made, unpurified horse chestnut paste over six weeks developed fluid around the heart and lungs that progressed to cardiac tamponade (compression of the heart by the fluid). After other causes were investigated and excluded, the case authors attributed it to the paste. The EMA noted that the preparation was poorly described and that this case did not raise a safety concern for standardised extracts.
These are single case reports. They cannot establish cause with certainty, but they illustrate that home-made and unstandardised horse chestnut preparations sit outside the safety evidence from trials.
Esculin or Aescin: Which Is the Poison?
Two compounds with confusingly similar names are involved, and published sources do not describe them the same way.
- Esculin (aesculin) is a coumarin glycoside — a small ring-shaped plant compound with a sugar attached. A 2018 mouse study describes it as a coumarin derivative found in Aesculus hippocastanum, and a 2024 study names horse chestnut bark as one of its sources. The EMA bark report states the bark contains up to 7% coumarin derivatives, chiefly esculin and fraxin.
- Aescin (escin) is a mixture of saponins — much larger, soap-like molecules — and is the main active compound of the seed.
NCCIH attributes the toxicity of the raw seeds, bark, flowers and leaves to esculin. The National Capital Poison Center names the toxin as "saponin aesculin", combining the two terms. The EMA's 2020 seed report, citing a 1992 pharmacognosy reference, states that coumarin derivatives such as aesculin and fraxetin are present in other parts of the tree "but not in the seeds or the seed shell", and separately records that β-aescin, one of the saponin fractions, can break open red blood cells in laboratory tests. A 2023 review describes β-escin as having "significant cytotoxicity".
Put together, the sources point to the saponins as at least part of the raw seed's toxicity, with esculin the characteristic compound of the bark. What this page can state with confidence is narrower: the raw seed, bark, flowers and leaves are all described as poisonous by reliable sources, and standardised seed extracts are a processed, measured product that has its own, separate safety record.
Esculin in the Laboratory
Esculin has a second, quite different research life. In a 2018 study, esculin reduced liver damage, inflammation and oxidative stress in mice given an experimental liver toxin (lipopolysaccharide with D-galactosamine). In a 2024 study, it reduced chemically induced liver fibrosis in mice and blocked the activation of scar-forming liver cells in culture. The EMA bark report notes that the acute toxicity of isolated esculin and esculetin is low in animal testing.
There is no contradiction here. A purified compound given at a measured dose to a laboratory animal is a different exposure from eating a seed, bark or leaf that contains a mixture of compounds. These mouse studies do not make raw horse chestnut safe to eat, and they are not evidence of a liver benefit in people.
Why the Extract Is a Different Story
The standardised seed extracts used in clinical trials are dried alcohol-and-water extracts measured for aescin content. Their safety record comes from controlled trials:
- The 2012 Cochrane review found adverse events usually mild and infrequent.
- A 2002 meta-analysis of 13 randomised trials and 3 observational studies (10,725 patients) found no severe adverse events, and no significant increase in mild ones.
- The EMA lists gastrointestinal complaints, headache, dizziness, itching and allergic reactions as the undesirable effects of oral use.
The extract has its own rare serious reports: LiverTox rates it a possible, rare cause of liver injury, and the EMA report records one case of liver cirrhosis after five years of use. Those reports concern the extract, not raw seeds, and are covered on the main page under safety.
The Coumarin and Bleeding Question
Because the tree contains coumarin compounds, horse chestnut often appears on lists of herbs that might interact with blood thinners. A 2002 review groups horse chestnut with chamomile, fenugreek and red clover as coumarin-containing herbs that could, in theory, add to bleeding risk with aspirin-type painkillers, while noting that most such interaction information is based on case reports, animal studies and test-tube data. The EMA seed report states that aescin has been said to increase the effect of anticoagulants, but that no confirmed case reports were found. The EMA bark report likewise describes the warfarin interaction as a theoretical risk with no direct experimental or clinical evidence.
Natural plant coumarins are not the same thing as the drug warfarin. For the clotting test that monitors warfarin, see PT/INR; for another coumarin story, see Cinnamon: Cassia vs Ceylon and Coumarin.
Key Facts at a Glance
- Raw horse chestnut seeds, bark, flowers and leaves are described as poisonous by NCCIH, the US poison-control service and the European forest atlas.
- Most raw-seed ingestions cause stomach upset; rare severe reactions include collapse, low blood pressure and airway swelling.
- Boiling did not prevent symptoms in a reported case.
- Horse chestnuts are round, bitter, usually one per husk; sweet chestnuts are pointed, starchy, several per furry husk.
- Sources disagree on whether esculin or aescin saponins drive raw-seed toxicity; the EMA reports esculin is absent from the seed itself.
- Standardised seed extracts have a separate safety record from clinical trials.
Anyone who has eaten part of a horse chestnut can get free, confidential advice from a poison center; in the US the number is 1-800-222-1222.
Key Research Papers
- Liu A, Shen Y, Du Y, et al.. Esculin prevents Lipopolysaccharide/D-Galactosamine-induced acute liver injury in mice. Microbial pathogenesis. 2018;125():418-422. PubMed PMID: 30290266
- Xu S, Chen Y, Miao J, et al.. Esculin inhibits hepatic stellate cell activation and CCl(4)-induced liver fibrosis by activating the Nrf2/GPX4 signaling pathway. Phytomedicine : international journal of phytotherapy and phytopharmacology. 2024;128():155465. PubMed PMID: 38471319
- Wang Y, Han X, Wan X, et al.. β-Escin: An Updated Review of Its Analysis, Pharmacology, Pharmacokinetics, and Toxicity. The American journal of Chinese medicine. 2023;51(8):2095-2120. PubMed PMID: 37865870
- Liang J, Jin W, Li H, et al.. In Vivo Cardiotoxicity Induced by Sodium Aescinate in Zebrafish Larvae. Molecules (Basel, Switzerland). 2016;21(3):190. PubMed PMID: 26907249
- Abebe W. Herbal medication: potential for adverse interactions with analgesic drugs. Journal of clinical pharmacy and therapeutics. 2002;27(6):391-401. PubMed PMID: 12472978
- Pittler MH, Ernst E. Horse chestnut seed extract for chronic venous insufficiency. The Cochrane database of systematic reviews. 2012;11(11):CD003230. PubMed PMID: 23152216
- Siebert U, Brach M, Sroczynski G, et al.. Efficacy, routine effectiveness, and safety of horsechestnut seed extract in the treatment of chronic venous insufficiency. A meta-analysis of randomized controlled trials and large observational studies. International angiology : a journal of the International Union of Angiology. 2002;21(4):305-15. PubMed PMID: 12518108
- Sirtori CR. Aescin: pharmacology, pharmacokinetics and therapeutic profile. Pharmacological research. 2001;44(3):183-93. PubMed PMID: 11529685
- Gallelli L. Escin: a review of its anti-edematous, anti-inflammatory, and venotonic properties. Drug design, development and therapy. 2019;13():3425-3437. PubMed PMID: 31631970
PubMed Topic Searches
- PubMed: Aesculus hippocastanum poisoning
- PubMed: esculin toxicity
- PubMed: horse chestnut seed intoxication
Other Sources
Connections
- Heart, Liver and Metabolic Herbs — the parent category
- Horse Chestnut (Aesculus hippocastanum) — the main page
- Horse Chestnut: Benefits Deep Dive — the hub
- Aescin: How It Works — the saponin in detail
- Chronic Venous Insufficiency Evidence — what the extract was tested for
- Haemorrhoids, Bruising and Topical Use — the topical side
- Quinoa: Saponin Removal — how a food's bitter saponins are washed off
- Cinnamon: Coumarin — another coumarin story
- PT/INR — the clotting test behind blood-thinner monitoring
- Liver Injury from Supplements — the wider picture
- Acute Kidney Injury — the kidney concern with injected aescin