Ancient cannabis use: 2,800-year-old Siberian tomb hair holds earliest dated sign

Hair from a Scythian princely tomb in Tuva, southern Siberia, carries cannabinol, a breakdown product of THC, and the tomb held about 1,000 cannabis seeds dated to the early 8th century BC, about 300 years before the previous oldest chemical evidence of cannabis use.

Peer reviewed · Archaeological chemistry · One site · Hair result disputed

A team led by archaeologist Gino Caspari of the University of Bern and the Max Planck Institute of Geoanthropology reports in Science that people buried at Tunnug 1, a monumental Scythian tomb in the Republic of Tuva, Russia, were exposed to psychoactive cannabis about 2,800 years ago. Several hair samples from the central burial chamber gave a clear signal for cannabinol (CBN) and a weak one for THC. The authors call it the earliest direct evidence of cannabis use. An independent forensic toxicologist says the hair data cannot rule out contamination from the seeds buried alongside.

Evidence at a glance

Is this a big deal?

It is a solid step with one contested link. The seeds and their date are hard evidence, and finding cannabis compounds in the hair of people buried with them is rare in archaeology. Whether that hair records a living person’s use, rather than 28 centuries next to cannabis, is the question the next round of testing has to settle.

What scientists found

Tunnug 1 is a large burial mound in the Uyuk Valley of Tuva, near Russia’s border with Mongolia. It was built for high-ranking members of Scythian society, the horse-riding nomads of the Eurasian steppe. The team had already reported 18 horses arranged on the mound, and the Bern release says later digging turned up several dozen more sacrificed horses plus scattered human bones. Human sacrifice is considered likely but has not been confirmed.

Inside, the excavators recovered about 1,000 cannabis seeds, according to the university. Keystone-SDA, the Swiss news agency, put the count at more than 1,400 intact seeds plus fragments; SciQuest could not check the paper’s own tally. The seeds probably came from wild plants, and the release says it is not possible to tell whether anyone grew them for their psychoactive properties.

The chemistry came from hair. Several samples from the central burial chamber gave a clear signal consistent with cannabinol, or CBN, which forms partly as THC (tetrahydrocannabinol, the main intoxicating compound) oxidises. A weaker signal matched THC itself by mass.

How they found it

The excavation began in 2017, run by a team from Germany, Russia, Sweden and Switzerland, AP reported. Dating was the first job. Six individual seeds went for radiocarbon dating, and together with other dates from the site they place the seeds in the early eighth century BC. Keystone-SDA reported that a statistical model narrows it to the first decade of that century. From 2026, the early 700s BC is roughly 2,790 to 2,830 years ago, which is where the “2,800 years” headline comes from.

The hair was the harder part. The team ran it through high-resolution mass spectrometry, which sorts molecules by mass precisely enough to pick out specific compounds. Think of CBN in old hair like a tide line on a beach: it tells you the water was there, not how high it came or exactly when. Where the comparison breaks down matters here: hair can also absorb compounds from outside, which is the heart of the dispute below.

Caspari told AP that “It’s extremely rare that we have both preserved plant parts and human soft tissue with the active compounds in it in the same context.” The release quotes him on the seeds: “The seeds show that cannabis played an important role in the rituals surrounding the burial.”

Why it matters to you

This won’t change anything about cannabis today, and it says nothing about its safety or medical use. It pins a date on one of the oldest relationships between people and a mind-altering plant, using the body of someone who lived it rather than a pot or a written account.

What the evidence does not show

It does not show how much anyone took, or how. The release says the hair cannot reveal dose or route. Later sites in Central Asia with cannabinoid residues on heated stones and inside vessels point to inhaling smoke, but that inference comes from other places, not this tomb.

It does not show strong plants. Nobody knows how much THC this cannabis carried, or whether people picked out stronger wild stands. The ritual link is also an interpretation. Seeds in a grave with sacrificed horses fit a funeral rite, but cannabis was used for seed oil and fibre too, as AP noted, so seeds alone would not prove drug use.

How this fits what we already knew

The previous benchmark came from the Jirzankal Cemetery in the eastern Pamirs of western China. In 2019, Ren and colleagues, with Robert Spengler, now at the Max Planck Institute of Geoanthropology, among the authors, reported cannabis compounds in wooden braziers from mortuary ceremonies around 500 BCE. Their abstract concludes that cannabis was smoked in ritual there by at least 2,500 years ago and that the plants produced high levels of psychoactive compounds.

Jirzankal residues came from burned equipment; Tunnug 1 adds hair from human remains, which is closer to the person. The Greek historian Herodotus, writing in the fifth century BC, described Scythians inhaling the fumes of burning hemp seed.

What an independent toxicologist says

Volker Auwärter, who heads forensic toxicology at the University Medical Center Freiburg in Germany and was not part of the study, told AP by email that the hair analysis “is scientifically weak”. In his view, the data cannot distinguish consumption from external exposure, since the hair could have picked up cannabis traces from being buried with the seeds. “For a central part of the study, the article falls far short of what would have been possible,” he wrote, adding that a forensic toxicology test would have been easy to arrange.

AP reported that some other independent researchers were not convinced either. SciQuest has not yet reached an outside archaeologist for comment and will add one if it does.

How other headlines framed it

• Associated Press (via CTV News) hedged its headline with “study suggests” and gave space to the toxicologist’s criticism.
• Popular Science called the tomb 2,700 years old and put “oldest evidence” in its headline without qualification.
• ScienceAlert framed it as the earliest known evidence of intoxication and quoted only the study team.
• Keystone-SDA (via blue News) said cannabis has been used for 2,800 years and noted that accidental ingestion cannot be fully ruled out.

The 2,700 and 2,800 figures describe the same date: the early 8th century BC is about 2,800 years ago, and the difference is rounding.

What happens next

The team has started sequencing DNA from the seeds, and Caspari says “DNA preservation is excellent in this case.” That could show whether these plants were wild or on the way to domestication. The other test is the one Auwärter asked for: forensic-style hair analysis that separates compounds inside the hair shaft from those on its surface. Until that is done, the consumption claim rests on the authors’ reading of the combined evidence. For more on what hair and residues can and cannot tell archaeologists, see our report on molecular traces of Egyptian embalming materials.

In one sentence

Seeds and hair from a Scythian tomb in Tuva put cannabis use at about 2,800 years ago, the oldest dated chemical sign yet, but whether the hair records a high or contamination is still argued.

Read next

• Ancient Egyptian embalming materials hold aloe traces: another case of chemistry reading what ancient people used.
• Eland collagen found in 60,000-year-old ochre: a residue whose purpose, like the tomb’s cannabis, is still debated.
• Ancient genomes link Celtic and Germanic spread to migrations: how DNA tracks movement, as the Tunnug seed sequencing aims to do.

Sources and disclosure

1. Caspari G., Sadykov T., Fomicheva D., Blochin J., Armbruster D., Happel O., Ryabogina N., Boxleitner K., Spengler R. “Direct evidence for psychoactive cannabis use 2800 years ago.” Science 394 (6824): 649–653 (2026). doi:10.1126/science.aeg3800.
2. University of Bern, 5 November 2026: Cannabis was used as an intoxicant as early as 2,800 years ago.
3. Ren M. et al. “The origins of cannabis smoking: Chemical residue evidence from the first millennium BCE in the Pamirs.” Science Advances 5 (6): eaaw1391 (2019). doi:10.1126/sciadv.aaw1391.
4. Associated Press, 5 November 2026: Seeds, preserved hair in tomb point to oldest signs of cannabis high 2,800 years ago, study suggests (full text read via La Nación’s reprint).
5. Coverage read for framing: Popular Science, ScienceAlert, Keystone-SDA via blue News.

Disclosure: This article was drafted with AI assistance and checked against the sources above. The full text of the Science paper was behind a paywall; SciQuest read its abstract, author list and affiliations through Crossref and relied on the University of Bern release for methods and results, so figures such as the number of hair samples are not given. Calculation by SciQuest: 2026 + 765 to 800 = about 2,790 to 2,830 years; Jirzankal (about 500 BCE) is about 2,525 years ago, roughly 300 years later. The graphic is a SciQuest original; no third-party images are used. Spotted an error? See our corrections policy.