Mugwort Essential Oil: Thujone Chemistry and Real Toxicity Data

Ask "how much thujone is in mugwort" and every source gives a different number — because there is no single number, and reaching for wormwood's well-known figures to fill the gap would be borrowing the wrong species' chemistry. The real, verified answer is more interesting than any single percentage: Artemisia vulgaris essential oil is chemically unstable across geography, sometimes barely containing thujone at all, and a real human poisoning case exists for this species specifically — not just for its more famous relative, wormwood.

Table of Contents

  1. The Chemotype Problem: Why a Single Percentage Is the Wrong Question
  2. Side by Side: Mugwort Essential Oil Is Not Wormwood Essential Oil
  3. When Thujone Does Dominate: The Geographic Range
  4. A Real Human Poisoning Case — From Tea, Not Oil
  5. Thujone's Mechanism: GABA-A Antagonism and Metabolic Detoxification
  6. Other Toxicity Signals: Brine Shrimp and Genotoxicity Screening
  7. Beyond Thujone: Real Antioxidant, Anti-Inflammatory, and Antimicrobial Findings
  8. Practical Safety Guidance
  9. Key Research Papers
  10. Connections

The Chemotype Problem: Why a Single Percentage Is the Wrong Question

The single most useful, most verifiable fact about mugwort essential oil chemistry is also the least satisfying to state as a headline: it varies enormously, plant to plant and region to region, and thujone is often not even the majority constituent. A 2016 survey of essential oils from fifteen wild Artemisia vulgaris populations across Lithuania found the two most common dominant compounds were davanones (13.8–45.5% in six of the fifteen oils) and germacrene D (9.1–30.5% in four oils) — trans-thujone appeared as a major component in only two of the fifteen populations tested, at 8.9% and 10.9%.[2] In thirteen of fifteen wild populations from a single country, in other words, thujone was not the dominant compound in the oil at all.

This variability is the honest headline, and it has a real practical consequence: any single number quoted for "mugwort's thujone content" is describing one population, grown in one place, harvested at one time — not the species. A Nepal-sourced sample analyzed by Pandey and colleagues in 2017 found β-thujone at 19.19%, alongside sabinene (11.29%), camphor (11.89%), and germacrene D (8.42%)[4] — genuinely thujone-substantial, and genuinely different from the Lithuanian profile above. Neither number is "the" answer; both are real, verified, species-locked measurements of real chemotype diversity.

Back to Table of Contents

Side by Side: Mugwort Essential Oil Is Not Wormwood Essential Oil

The most direct evidence against borrowing wormwood's thujone reputation for mugwort comes from a 2025 Serbian study that did something unusually useful: it analyzed Artemisia absinthium (wormwood) and Artemisia vulgaris (mugwort) essential oils side by side, from the same region, using the same methodology.[3] The wormwood oil was dominated by β-thujone at 18.9%, consistent with its reputation as the thujone-rich Artemisia. The mugwort oil's dominant constituents were gurjunene (10.41%), cis-chrysanthenyl acetate (7.17%), and γ-humulene (6.67%) — no thujone at all among the leading compounds in this particular Serbian population. The same study found mugwort's plant extract had higher antioxidant activity than wormwood's (IC50 0.171 mg/mL versus 0.263 mg/mL — a lower IC50 means a smaller dose achieves the same antioxidant effect, so mugwort was the more potent of the two here), while wormwood's extract was the stronger weed-suppressant.

This is a direct, same-study, same-methodology demonstration that the two species are genuinely chemically distinct populations of the same genus, not interchangeable sources of the same essential oil — exactly the kind of comparison this site's evidence doctrine asks for before letting one Artemisia species' well-known chemistry stand in for another's. A separate GC-MS survey comparing four Artemisia species — tarragon, southernwood, wormwood, and mugwort — for antifungal activity against Candida albicans reached a parallel conclusion from a different angle: the strongest antifungal activity in that comparison belonged to southernwood, not mugwort, which is worth stating plainly rather than letting a multi-species screening study get summarized as "Artemisia is antifungal" in a way that quietly credits mugwort with a result that was actually strongest in a different species entirely.[11]

Back to Table of Contents

When Thujone Does Dominate: The Geographic Range

None of this means mugwort essential oil is thujone-free or that thujone concerns are misplaced — only that a single number is the wrong way to describe it. Older Indian material, reviewed in secondary literature discussing populations from the Nilgiri Hills, has been reported with α-thujone as high as the 40–60% range in some samples, which would be genuinely wormwood-comparable or higher; the Nepal sample above sits at a substantial-but-more-moderate 19%; several Lithuanian and the Serbian populations above show little to none. The honest, defensible statement this page can stand behind is a range, not a figure: published measurements of thujone in Artemisia vulgaris essential oil span from essentially absent to the majority constituent, depending on where and when the plant was grown, and any product not chemically tested should be assumed to fall anywhere in that range.

Back to Table of Contents

A Real Human Poisoning Case — From Tea, Not Oil

This is the single most important safety citation on this page, and it is worth reading closely because of what it explicitly says about itself. In 2018, Di Lorenzo and colleagues published a case report in the Journal of Food Science describing a 49-year-old man admitted to a Zurich emergency room in a manic state after drinking roughly one liter of an Artemisia vulgaris infusion (tea).[1] Using validated HPLC and gas-chromatography methods, the authors measured serum thujone directly: 27.7 ± 3.48 µg/mL on the day of admission, still 24.1 ± 0.15 µg/mL the following day — confirmed thujone poisoning from an oral tea preparation, not from concentrated essential oil.

The authors' own framing of why this case mattered enough to publish is directly relevant to this page's central theme: they note that adverse thujone effects had previously been documented only for essential oil from Artemisia absinthium (wormwood), and that this case was one of the first to confirm the same poisoning mechanism from Artemisia vulgaris specifically, consumed as an infusion. In other words, the scientific literature itself had the exact gap this site's evidence doctrine warns about — wormwood's thujone toxicity data standing in for mugwort's, by default, until someone checked. The authors offer four possible explanations for how a tea produced serum levels that high: an unusually thujone-rich source plant, chronic/repeated exposure rather than a single incident, individual variation in how efficiently the patient metabolized thujone, or — notably — contamination or adulteration of the plant material with a different, more thujone-rich Artemisia species such as wormwood itself. That last possibility is a real acknowledgment, from within the case report, that even a confirmed vulgaris-labeled poisoning cannot fully rule out a supply-chain identity problem, which is a useful, honest note of humility to carry into any discussion of dried herbal material bought without species verification.

Back to Table of Contents

Thujone's Mechanism: GABA-A Antagonism and Metabolic Detoxification

Thujone's mechanism of harm is well characterized at the molecular level, in research conducted independently of which Artemisia species it was sourced from — a genuinely useful distinction, because the chemistry of the isolated molecule does not change depending on its plant of origin, even though its concentration in any given plant does. The landmark study, by Höld and colleagues in 2000, established that alpha-thujone acts as a GABA-A receptor antagonist — it blocks the same inhibitory chloride-channel receptor that benzodiazepines and barbiturates enhance — which is the mechanistic basis for thujone's convulsant potential at sufficient dose.[7] That paper studied thujone via wormwood/absinthe, so it is cited here specifically and only for what thujone the compound does, not as mugwort-specific evidence; the concentration data earlier on this page is what supplies the mugwort-specific half of the picture.

A companion study by an overlapping author group examined how the body detoxifies thujone via cytochrome P450 metabolism, finding site-specific and species-specific differences in how efficiently alpha- and beta-thujone are cleared.[8] This matters practically: the Zurich poisoning case above raised individual metabolic variation as one candidate explanation for an unexpectedly severe reaction to what should have been a modest dose, and this detoxification research is the mechanistic backdrop for why that is a biologically reasonable possibility rather than a hand-wave.

Back to Table of Contents

Other Toxicity Signals: Brine Shrimp and Genotoxicity Screening

Beyond thujone specifically, a small amount of direct Artemisia vulgaris essential-oil toxicity screening exists and is worth reporting rather than assuming. The same Lithuanian chemotype study cited above also tested its fifteen essential-oil samples for toxicity using a brine shrimp (Artemia sp.) lethality assay, a standard, inexpensive preliminary toxicity screen — not a substitute for mammalian or human data, but a real first-pass signal. Oils richer in germacrene D, cineole, camphor, and davanone showed measurable toxicity in this assay, with LC0 values (the highest concentration causing no lethality) of 10.3–23.1 µg/mL after 24 hours.[2] Separately, a 2022 comparative study directly tested the genotoxic activity of Artemisia vulgaris extract against a related species, Artemisia alba, in vitro — real, species-locked genotoxicity screening data, distinct from and additional to the thujone-specific neurotoxicity concern above.[9] Neither finding translates directly into a human dietary risk at typical culinary or occasional-tea exposure levels, and this page is not going to manufacture that leap; they are reported because a preliminary toxicity signal is still a real finding, and omitting it would understate what is actually known.

Back to Table of Contents

Beyond Thujone: Real Antioxidant, Anti-Inflammatory, and Antimicrobial Findings

Thujone and toxicity are not the whole chemistry story, and it would be its own kind of imbalance to let a page about mugwort's essential oil and extract chemistry read as purely cautionary. Several recent, real, species-locked studies document genuine bioactivity beyond the thujone question. A 2024 Vietnamese study found that ethanol extracts of Artemisia vulgaris produced significant anti-inflammatory activity in the standard carrageenan-induced paw edema model in mice, alongside real antioxidant activity (DPPH, ABTS, and hydrogen-peroxide assays) and inhibition of xanthine oxidase — the same enzyme allopurinol targets to lower uric acid — with five isolated flavonoids (luteolin, rutin, apigenin, myricetin, and quercetin) identified as the likely active constituents, consistent with the plant's traditional Vietnamese use for gout and arthritis.[5] A newly identified compound named Artemvulactone E, isolated from Artemisia vulgaris in 2024, showed genuine anti-inflammatory activity against lipopolysaccharide-induced inflammation in both a macrophage cell line and a zebrafish model — two different preclinical systems agreeing.[6] A separate 2025 rat study confirmed anti-inflammatory activity in the same carrageenan paw-edema model using whole extract alongside the isolated compound β-caryophyllene oxide.[10]

Antimicrobial data is similarly real, if modest and mostly preclinical. The Nepal essential-oil study above also found meaningful antioxidant and antibacterial activity, with the methanol extract outperforming the essential oil itself on both measures.[4] A 2025 study screened methanolic Artemisia vulgaris extract against Bacillus cereus, a foodborne pathogen associated with reheated and improperly stored cooked food, and found a genuine inhibitory effect.[12] A 2026 study testing extracts against the parasites Leishmania major and Trichomonas vaginalis found real, dose-dependent, selective (non-cytotoxic) antiprotozoal activity for Artemisia vulgaris, though a comparison extract from Berberis vulgaris (barberry) performed somewhat better on the same assays in the same study.[13] None of this is clinical-trial-grade evidence of a treatable human condition; all of it is real, recent, species-locked laboratory data consistent with mugwort being a genuinely bioactive plant whose chemistry deserves more than a one-line "contains thujone" summary.

Back to Table of Contents

Practical Safety Guidance

Put together, the chemistry above supports a few concrete, defensible practical points. Occasional culinary use — the amount of mugwort leaf actually eaten in a serving of Japanese yomogi rice cakes or Korean ssuk soup, as covered on the main Mugwort page — is a small fraction of the plant material involved in the Zurich poisoning case (a full liter of concentrated infusion) and is not comparable to it. Regularly brewed, strong mugwort tea consumed in quantity and repeatedly is a materially different exposure, and the poisoning case above is the direct evidence for why that distinction matters, not a hypothetical one. The concentrated essential oil should never be taken internally under any circumstances, regardless of a given bottle's specific thujone percentage, because that percentage is never disclosed on a typical consumer product and the chemotype range above shows it cannot safely be assumed. And because no consumer-grade dried mugwort or tea product on the market is chemically tested and labeled for thujone content the way the research studies above were, this page cannot respons­ibly convert any of this chemistry into a specific "safe" dose or cup-per-day figure — the honest answer, per this site's evidence doctrine, is to state that range and its uncertainty rather than invent a number precise enough to sound authoritative.

Back to Table of Contents

Key Research Papers

  1. Di Lorenzo C, Ferretti F, Moro E, et al. Identification and Quantification of Thujone in a Case of Poisoning Due to Repeated Ingestion of an Infusion of Artemisia Vulgaris L. Journal of Food Science. 2018;83(8):2257-2264. — PubMed: Artemisia vulgaris thujone poisoning — the core human case report; confirmed serum thujone from tea, not oil.
  2. Judzentiene A, Garjonyte R. Compositional Variability and Toxic Activity of Mugwort (Artemisia vulgaris) Essential Oils. Natural Product Communications. 2016;11(9):1353-1356. — PubMed: Artemisia vulgaris Lithuania toxic — fifteen-population chemotype survey; thujone dominant in only 2 of 15 oils.
  3. Tojić T, Đorđević T, Đurović-Pejčev R, et al. Allelopathic Potential of Artemisia absinthium and Artemisia vulgaris from Serbia: Chemical Composition and Bioactivity on Weeds. Plants. 2025;14(11):1663. — PubMed: Artemisia absinthium vulgaris Serbia — direct same-study, same-methodology species comparison; mugwort oil not thujone-dominant.
  4. Pandey BP, Thapa R, Upreti A. Chemical composition, antioxidant and antibacterial activities of essential oil and methanol extract of Artemisia vulgaris and Gaultheria fragrantissima collected from Nepal. Asian Pacific Journal of Tropical Medicine. 2017;10(10):952-959. — PubMed: Artemisia vulgaris Nepal essential oil — Nepal chemotype, thujone 19.19%; real antioxidant/antibacterial data.
  5. Trinh PTN, Truc NC, Danh TT, et al. A study on the antioxidant, anti-inflammatory, and xanthine oxidase inhibitory activity of the Artemisia vulgaris L. extract and its fractions. Journal of Ethnopharmacology. 2024;334:118519. — PubMed: Artemisia vulgaris xanthine oxidase — in vivo anti-inflammatory confirmation; five active flavonoids identified.
  6. Artemvulactone E isolated from Artemisia vulgaris L. ameliorates lipopolysaccharide-induced inflammation in both RAW264.7 and zebrafish model. Frontiers in Pharmacology. 2024. — PubMed: Artemvulactone — newly identified vulgaris-specific compound, anti-inflammatory in two model systems.
  7. Höld KM, Sirisoma NS, Ikeda T, Narahashi T, Casida JE. Alpha-thujone (the active component of absinthe): gamma-aminobutyric acid type A receptor modulation and metabolic detoxification. Proceedings of the National Academy of Sciences. 2000;97(8):3826-3831. — PubMed: thujone GABA absinthe — landmark thujone mechanism paper; compound-level evidence, studied via wormwood.
  8. Detoxification of alpha- and beta-Thujones (the active ingredients of absinthe): site specificity and species differences in cytochrome P450 oxidation in vitro and in vivo. Chemical Research in Toxicology. 2001. — PubMed: thujones cytochrome P450 detoxification — companion metabolism study; individual variation in thujone clearance.
  9. Comparative study of the genotoxic activity of Artemisia vulgaris L. and Artemisia alba Turra extracts in vitro. Drug and Chemical Toxicology. 2022. — PubMed: Artemisia vulgaris genotoxic comparative — direct genotoxicity screening, species-locked.
  10. In vivo and in silico anti-inflammatory activity of Artemisia vulgaris and β-caryophyllene oxide in carrageenan-induced paw edema in Wistar rats. Drug and Chemical Toxicology. 2025. — PubMed: Artemisia vulgaris carrageenan caryophyllene — second independent carrageenan-model anti-inflammatory confirmation.
  11. Obistioiu D, Cristina RT, Schmerold I, et al. Chemical characterization by GC-MS and in vitro activity against Candida albicans of volatile fractions prepared from Artemisia dracunculus, Artemisia abrotanum, Artemisia absinthium and Artemisia vulgaris. Chemistry Central Journal. 2014;8(1):6. — PubMed: four Artemisia species Candida — four-species comparison; strongest antifungal activity belonged to southernwood, not mugwort.
  12. Fareid MA, El-Sherbiny GM, Elafandy NM, Eltoum NE. Assessment of the risk associated with Bacillus cereus isolates and potential combat with methanolic Artemisia vulgaris extract. Frontiers in Microbiology. 2025;16:1640200. — PubMed: Bacillus cereus Artemisia vulgaris — real inhibitory activity against a foodborne pathogen.
  13. Aksoy T, Girginkardeşler N, Balcıoğlu İC, Kilimcioğlu AA. Antiprotozoal Activity of Artemisia vulgaris and Berberis vulgaris Against Leishmania major and Trichomonas vaginalis. Acta Parasitologica. 2026;71(1):44. — PubMed: Artemisia vulgaris Leishmania Trichomonas — selective, dose-dependent antiprotozoal activity.

Back to Table of Contents

Connections

Back to Table of Contents