Material guide: tea tree oil — the essential oil with the most published allergy cases, and a dermal toxicity field containing one line about a rabbit

About 6 minutes.

Short version

• Name: tea tree oil, from Melaleuca alternifolia, CAS 8022-72-8
• Database category: antimicrobial agents — not fragrance
• cosmetic_uses: antioxidants; perfuming agents
• Odour type: spicy; odour: citrus, spicy, terpenic, nutmeg, pine
• Strength: medium; substantivity 44 hours; flash point 50 °C
• Optical rotation: +5.00 to +15.00
• Suppliers: 101; FEMA 3902
• GHS: H302, H313, H315 skin irritation, H317 skin sensitisation Category 1, H319, H411, H226

It is not classified as a fragrance material

The category field reads antimicrobial agents.

The notes field explains why:

…essential oil extracted from Melaleuca alternifolia (tea tree). it is used as a topical antimicrobial due to the presence of terpineol.

This material's principal identity in the database is an antimicrobial; the fragrance use is incidental. (Same as pine oil, whose principal identity is a household-product feedstock.)

Its odour field does have content: citrus, spicy, terpenic, nutmeg, pine, at 44 hours.

It is the allergy champion among essential oils

De Groot and Schmidt devoted a review to it in Contact Dermatitis in 2016, and their statement is direct:

Of all essential oils, TTO has caused most (published) allergic reactions since the first cases were reported in 1991. In routine testing, prevalences of positive patch test reactions have ranged from 0.1% to 3.5%. Nearly 100 allergic patients have been described in case reports and case series.

They also list commercial tea tree oil's major constituents:

terpinen-4-ol, γ-terpinene, 1,8-cineole, α-terpinene, α-terpineol, p-cymene, and α-pinene.

And then the key sentence:

Fresh TTO is a weak to moderate sensitizer, but oxidation increases its allergenic potency. The major sensitizers appear to be ascaridole, terpinolene, α-terpinene, 1,2,4-trihydroxymenthane, α-phellandrene, and limonene.

Look at the structure of that: of the major constituents and the major sensitisers, only α-terpinene appears on both lists.

What causes the allergy is mostly not what was in it, but what it turned into.

(The same pattern appears with linalool, whose sensitisation also rises on oxidation, and the general case is Storage and oxidation.)

Ascaridole's dose–response

Christoffers and colleagues addressed a very concrete question in the same journal in 2014: what patch-test concentration should ascaridole be tested at?

They patch tested 319 eczema patients with ascaridole at 1%, 2% and 5%, and 250 patients with oxidised tea tree oil at 5%, reading on day 3 and day 7.

The results:

Ascaridole concentration	Positive
1%	1.4%
2%	5.5%
5%	7.2%

Raise the concentration, raise the positives.

But the paper does not stop there. They also recorded:

However, the frequencies of irritant and doubtful reactions also increased, especially for ascaridole 5%.

That is the most valuable part of the paper. Looking only at positivity, 5% is the best test concentration; but that 5% also produces more reactions you cannot tell apart from irritation. Catching more and catching correctly are different things.

They also confirmed that a positive to ascaridole is related to a positive to tea tree oil, supporting ascaridole as a sensitiser.

And the database's dermal toxicity field is one line

All of the above — allergy champion, oxidation raising potency, six major sensitisers, a dose–response curve — appears in this database's toxicity_dermal field as:

skin-rabbit LDLo 5000 mg/kg (Food and Chemical Toxicology, 1988)

One line, one rabbit, one lethal dose.

Its GHS field, meanwhile, carries both H315 (skin irritation, Category 2) and H317 (skin sensitisation, Category 1).

One material, two fields, entirely different amounts of information. That is exactly the problem measured in 94.5% of the dermal toxicity field is a rabbit's lethal dose — the field's name suggests it answers dermal questions.

Its oral field carries two child entries

The toxicity_oral field holds three entries, the last two at the same dose with different effects:

oral-rat LD50 1900 mg/kg (1988)

oral-child TDLo 500 µL/kg: BEHAVIORAL: "HALLUCINATIONS, DISTORTED PERCEPTIONS"; BEHAVIORAL: ATAXIA (J Toxicol Clin Toxicol, 1994)

oral-child TDLo 500 µL/kg: BEHAVIORAL: SLEEP; BEHAVIORAL: ATAXIA (Vet Hum Toxicol)

Two independent sources, the same dose, both children.

Tea tree oil is the best-supplied entry (101) among the 156 records carrying human toxicity data I collected last time.

This is an ingestion scenario, not a use scenario. But it explains why this material is worth labelling and storing carefully — it is common, cheap, present in many houses, and it has two recorded child poisonings.

How to use it

• Buy fresh and finish it. Oxidation raises sensitising potency, and its shelf life is 24 months. Dark, sealed, cool.
• Check IFRA before anything leave-on. GHS gives it skin sensitisation Category 1.
• Its odour direction is terpenic, medicinal, pine and nutmeg, not citrus, despite citrus being the first word in its odour field.
• Its flash point is 50 °C, in the band that matters for shipping (Shipping a flammable liquid).
• It is an antimicrobial, not a fragrance material. If you want the smell rather than the function, consider terpinen-4-ol or cineole on their own.

What this doesn't establish

• I have not smelled it. The whole article is fields, database-wide counts and two papers.
• "0.1% to 3.5%" is positivity in patch-tested populations — people tested because contact dermatitis was suspected, not a general-population rate.
• De Groot's paper is a review; I quote its statements and lists without going back to the cases it collates.
• I cited another series by this author in Forty-six records say keep it off your skin. Same author, different paper; this one is the tea tree monograph.
• Christoffers tested ascaridole, not tea tree oil. The 1.4% → 5.5% → 7.2% figures are ascaridole positivity at three test concentrations, not a sensitisation rate for the oil.
• That study's subjects were 319 eczema patients, a high-risk population by construction.
• Degree of oxidation cannot be read off the fields. The database records neither sample freshness nor peroxide value.
• I did not check the current IFRA standard for this oil.