What does human hair reveal about ancient cannabis use?

A Hair, a Grave, and an Unexpected Chemical Confession

A strand of human hair, nearly three millennia old, still carries a secret. Preserved inside a burial mound on the Siberian steppe, that hair held a molecular trace that no ancient text could have provided: the chemical fingerprint of a cannabis high.

The finding comes from Tunnug 1, an early Iron Age royal burial mound in the Uyuk Valley of Siberia, and it is not simply another old seed in an old grave. Archaeologist Gino Caspari at the University of Bern and the Max Planck Institute for Geoanthropology led the study, published in the journal Science in November 2026. His team identified tetrahydrocannabinol - THC, the molecule responsible for cannabis intoxication - directly in preserved human hair. That matters enormously, because hair does not just touch a substance. Hair metabolizes it. It records consumption from the inside.

Here is the strange part. For decades, archaeologists argued about ancient cannabis the way detectives argue about a weapon found near a body but not in anyone's hand. Seeds could mean fiber. Charred remains could mean accidental burning. But metabolites locked in hair tissue cannot mean anything except that this person absorbed THC into their bloodstream. The body kept the receipt.

That receipt is now the oldest of its kind. Researchers confirmed the Tunnug 1 find pushes the direct chemical evidence of intentional cannabis intoxication back 2,800 years - approximately 300 years further than any previously confirmed record. People in the 8th century BCE, on the edge of the Eurasian steppe, were deliberately getting high. We just finally have the proof that could not be argued away.

How Do You Prove a 2,800-Year-Old High?

For most of archaeology's history, finding cannabis at a dig site meant finding seeds - scorched remnants in a hearth, scattered across a floor, suggestive but never conclusive. Seeds prove presence. They do not prove anyone was affected by what they burned.

That changes when you go smaller. Much smaller.

Human hair is, among other things, a chemical diary. As a strand grows, it incorporates compounds circulating in the bloodstream at the time - including the metabolic byproducts of whatever the person consumed. This is phytochemical analysis: reading those molecular traces long after the body is gone. A single strand, properly preserved, can carry a record of what was inside a person weeks before death. At Tunnug 1, Gino Caspari's team found hair old enough to have belonged to someone alive when Homer was dictating or inventing the Iliad, and that hair still held its chemical signature.

The specific signal they were reading for was tetrahydrocannabinol - THC. Worth noting: the cannabis plant did not produce THC for our benefit. It is a secondary metabolite, a compound evolved for the plant's own purposes, possibly deterring insects or pathogens. We simply discovered, at some point in deep prehistory, that it was interesting to us.

What made Tunnug 1 genuinely unprecedented was the combination — the first time plant remains and direct proof of human intake were recovered together from a single excavation.

The site yielded more than 1,400 cannabis seeds alongside the chemical evidence of consumption locked in human tissue. Presence and consumption. Both, at once, in the same frozen ground.

Inside Tunnug 1: A Royal Burial Mound Full of Seeds

The Uyuk Valley sits in what is now the Tuva Republic of southern Siberia, a broad steppe corridor that has watched nomadic peoples move through for millennia. Tunnug 1 is what archaeologists call a "royal mound" - a kurgan of unusual scale and construction, the kind reserved for someone whose death demanded ceremony. The designation matters. This was not a roadside grave. It was a statement.

Inside that statement, Gino Caspari's team from the University of Bern and the Max Planck Institute for Geoanthropology found more than 1,400 seeds of Cannabis ruderalis. Put that number into a room: imagine filling a small glass jar, seed by seed. That jar does not arrive in a grave by accident.

The seeds survived because Siberian permafrost is, in practical terms, a freezer that nobody switched off. Where organic material - hair, wood, plant tissue - rots to nothing in a temperate climate within decades, the cold holds structure intact across centuries. Archaeologists working in Siberia sometimes describe it as reading a letter that the sender forgot to seal. What permafrost preserved here was not a hint of cannabis but a stockpile.

What Caspari's team then did was connect two threads that had never before been pulled tight in one place: the physical seeds on one hand, the chemical signature of THC in human hair on the other. Prior discoveries had offered one or the other. Here, for the first time, both came from the same site, the same event, and - almost certainly - the same people.

What China Adds: A Burial Shroud Made of Plants and a Shaman's Stash

Picture a man laid out in the dry heat of the Turpan Basin, sometime around 800 BCE. He is roughly 35 years old. Someone has arranged thirteen cannabis plants across his body - roots still attached, seed clusters still present, the tallest stems stretching nearly 90 centimeters. Not a handful of seeds, not an accidental scatter. A deliberate green blanket, pulled over him like a final gesture.

The Jiayi cemetery, excavated in the Xinjiang region of western China, produced one of the most visually arresting finds in archaeobotany. Those intact roots mattered enormously to the researchers: they proved the plants were grown locally, not carried in from some distant supply. Someone in this community cultivated cannabis on purpose, with enough care and knowledge to harvest whole plants at the right moment and use them in ritual.

The man himself had Caucasian physical features, which places him among the Subeixi culture - a people of the Turpan Basin whose connections stretched far across the Central Eurasian steppe.

A separate site nearby tells a different story with equal precision. At Yanghai cemetery, excavations in 2008 uncovered a shaman's grave containing 789 grams of processed female cannabis flowers. Not scattered seeds. Processed flowers, the part of the plant carrying the highest concentration of active compounds, stored in a leather bag and a wooden bowl - the kind of careful, intentional storage you use for something valuable and purposeful.

The dates of Jiayi and Tunnug 1 in Siberia are, within the resolution of radiocarbon dating, essentially the same. That near-simultaneity carries a message. By the 8th century BCE, knowledge of what cannabis could do was not a local secret in one mountain valley or one desert basin. It was already distributed across Central Eurasia, carried - like so many ideas - along the steppe roads that connected these communities long before anyone wrote any of it down.

Ritual, Smoke, and the Hypothesis Nobody Finds Comfortable

Consider what the archaeologists are actually proposing. Not that someone in the Uyuk Valley relaxed with cannabis after a long week. Not that a healer used it to ease a fever. The hypothesis is colder than that: the plant may have been used to sedate people, and horses, before they were killed as sacrificial offerings at high-status burials.

That reading is uncomfortable. It is also hard to dismiss.

Human and horse sacrifice at Scythian and steppe nomad burials is not speculative - it is one of the better-documented features of Iron Age mortuary practice across the Eurasian steppe. The Greek historian Herodotus described Scythian smoke baths in enough detail that scholars have cited him for two and a half millennia. The problem was always that his account was just a description, words on papyrus with no chemical fingerprint behind them. Tunnug 1 now puts metabolites in human tissue where before there were only ancient sentences.

The idea that cannabis traveled the steppe as a ritual technology - not as a trade good, not as a food crop - is what the near-simultaneous dates across Siberia and western China keep suggesting. These cultures may never have met. The plant, used this specific way, reached both of them by around 800 BCE anyway. Whether that happened through contact, convergence, or some older shared ancestor of the practice, nobody can yet say. That is not a failure of the evidence. It is simply where the question now lives.

Three Things the 2,800-Year-Old Cannabis Evidence Still Cannot Answer

The hair strand recorded the molecule. It did not record the gesture. THC found in preserved hair proves the substance reached the bloodstream, but nothing in that chemical signal tells us whether these people inhaled smoke, drank an infusion, or ate the plant as food. The delivery method, which shaped the experience entirely, remains invisible to us.

The second gap is a sampling problem. Every site discussed here is a high-status burial, a royal mound, a shaman's grave, a man honored with thirteen cannabis plants laid across his body. The archaeological record disproportionately preserves the deaths of the important. Whether cannabis was woven into ordinary daily life across the steppe, or reserved strictly for the rituals of a select few, the common graves cannot tell us what the common graves do not contain.

The third question is genetic, and it cuts deepest. Over 1,400 Cannabis ruderalis seeds came out of Tunnug 1, but seeds alone cannot reveal how much selective pressure humans had already applied to the plant by 800 BCE. Were these people cultivating for potency, nudging the chemistry across generations? We genuinely do not know — and that question, sitting quietly at the edge of what the data can reach, may be the most important one remaining: not just what they used, but whether they were already, deliberately, changing it.