Ancient DMT: what snuff tubes, ritual bundles and residue chemistry really show

THE SHORT VERSION · SWIPE

▶ Watch it as a 20-second video

‹›

Online claims about “ancient DMT” are often attached to a single image of a carved snuff tray or a mummy. The archaeology underneath is real, but narrower than the headlines. Below are the best documented finds, what each one does and does not establish, and why researchers still argue about how to read a trace of a molecule on a thousand-year-old pouch.

What counts as evidence of ancient DMT use?

Archaeologists have four main kinds of evidence. Iconography: carvings and pottery that may depict plants or people in altered states. Paraphernalia: tubes, trays, spoons and pouches shaped like objects used for snuffing in recorded Indigenous practice. Plant remains: seeds or microscopic starch grains that can be identified to a plant. And chemical residues: molecules extracted from the surface of an object and identified with mass spectrometry. The Chavín team stresses that for centuries the first two were all anyone had at that site: neither psychoactive plant remains nor chemical traces had been identified there before their study (Rick et al., 2025). Evidence: established.

Most of the direct evidence involves vilca (also called cebil), a tree the excavation papers identify from plant remains. In those papers its chemical marker is bufotenine (5-hydroxy-DMT), not DMT; the Bolivian bundle authors describe bufotenine as its primary tryptamine (Miller et al., 2019). That matters: a find described as “ancient DMT” is often, chemically, mostly a bufotenine find. Evidence: established.

What was found at Chavín de Huántar?

Chavín de Huántar, in the north-central highlands of Peru, is a monumental temple complex of the first millennium BCE, long interpreted as a centre of public ritual. In 2017–2018 excavators opened Gallery 3, a small underground chamber about 7 metres long, which had been sealed from shortly after the Chavín period until excavation. Deposits linked to the gallery’s use were dated by radiocarbon and Bayesian modelling to roughly 800–350 BCE (Rick et al., 2025). Evidence: established.

The gallery held 23 artifacts of forms associated elsewhere with consuming psychoactive plants: 22 made of bone, mostly bird wing bones shaped into short tubes, and one shell. Each was tested by gas chromatography–mass spectrometry, and 11 were also examined under the microscope, by an analyst working blind to the chemical results (Rick et al., 2025). In summary:

• Six bone artifacts tested positive for at least one diagnostic alkaloid: nicotine, bufotenine or DMT.
• DMT was identified in one artifact, a broken bone semicylinder, together with bufotenine. Bufotenine was found in four artifacts and nicotine in five.
• Four of the six positive artifacts also held microscopic starch grains, crystals and plant hairs matching vilca and wild tobacco (Nicotiana).
• No combustion products of the alkaloids were detected, which the authors read as support for inhalation rather than smoking.
• Several tubes carried evidence of both plants, which the authors interpret as likely mixed vilca and tobacco powders, the oldest such evidence they know of in the Americas.

Each item above comes from the paper’s results and discussion (Rick et al., 2025). The authors call these finds the northernmost direct evidence of vilca and tobacco use in the pre-Hispanic Andes. Because most tubes were found in restricted-access galleries rather than middens, they argue that consumption was organized and perhaps controlled, part of institutional ritual rather than individual practice. Evidence: established for the chemistry and context; the social interpretation is the authors’ argument, and preliminary.

How other researchers responded

The paper drew two published responses. A commentary in the same journal calls the botanical identification groundbreaking, and notes an open question: similar bone tubes also turn up in household areas at Chavín, so it is unclear whether they were made by craft specialists for temple use or used in smaller domestic rituals (Nesbitt, 2025). Evidence: an open question raised in a commentary.

A letter from an archaeologist at Memorial University of Newfoundland asks why the two plants were combined. He suggests tobacco’s own enzyme-inhibiting compounds might have changed vilca’s effects, and argues this complicates claims of prehistoric ayahuasca that rest only on detecting harmala compounds (Whitridge, 2025). Evidence: the letter’s pharmacological idea is a hypothesis, untested on these artifacts.

What was in the 1,000-year-old Bolivian ritual bundle?

Cueva del Chileno is a rock shelter at about 3,890 metres in the Lípez highlands of southwestern Bolivia, excavated in 2008 and 2010. Among what the excavators call “ritual trash” was a well-preserved leather bag. Inside were two carved wooden snuffing tablets, a snuffing tube with human hair braids attached, two camelid-bone spatulas, a woven headband, two small pieces of plant material on strings, and a pouch made of three fox snouts stitched together. The leather was radiocarbon dated to about 905–1170 CE (Miller et al., 2019). Evidence: established.

The team scraped a tiny sample from inside the fox-snout pouch and took a surface sample from one plant fragment, then analysed both with liquid chromatography–tandem mass spectrometry. The pouch yielded cocaine and its breakdown product benzoylecgonine, harmine, bufotenine, DMT, and a peak that might correspond to psilocin. The plant fragment yielded cocaine, benzoylecgonine and bufotenine (Miller et al., 2019). Evidence: established for the detections; the psilocin peak was not in the original screen and is a possibility, not an identification.

How the authors read it is careful. They link cocaine to coca leaf and bufotenine to vilca seeds. They link harmine to the vine used in ayahuasca. They are explicit that DMT is confounding, because it occurs in small amounts in vilca and so could come from the same material as the bufotenine, or from a separate source. They conclude that ayahuasca, or a combined snuff, “must be considered”, but that the result is not conclusive evidence of ayahuasca as drunk today. No human remains were associated with the bundle, so there is no direct evidence of consumption (Miller et al., 2019). Evidence: established that harmine and DMT traces sat in one pouch; whether they were ever combined in a drink is unresolved.

How old is the Andean snuff tradition?

The southern Andes have the longest direct record. A 2024 review of vilca use by a researcher at Argentina’s national research council (CONICET) states that the earliest evidence, about 4,000 years old, is vilca seeds found with pipes in northwest Argentina, and that remains have since been found in pipes, enema syringes, leather bags, tubes and trays, all confined to the Andean area: northern Chile, northwest Argentina, central and southern Peru, and Bolivia (Lema, 2024). The Chavín paper, which shares that author, dates the combined presence of vilca and tobacco in Argentine pipes to as early as 2500 BCE (Rick et al., 2025). Evidence: established for the existence of the finds; exact earliest dates vary between sources.

One of the classic Argentine studies examined two mummified bodies from the Alero I (La Matanza) site in the Puna of Jujuy, found with snuffing equipment. Using light, petrographic and scanning electron microscopes, the team identified the powder in one of the tubes as vilca by comparison with reference material, confirming a link that had until then rested on ethnographic sources (Pochettino et al., 1999). Evidence: established.

Across the border in Chile, chemical and contextual analysis of snuff powders from pre-Hispanic San Pedro de Atacama in Chile was published in 1991 (Torres et al., 1991); the Chavín paper cites wooden snuff trays from San Pedro dated about 500–1500 CE as the only earlier evidence of mixed vilca and tobacco snuff (Rick et al., 2025). Evidence: established.

Human tissue adds a different kind of evidence. The Bolivian bundle paper summarises hair studies of mummies from the Azapa Valley in northern Chile: bufotenine in the hair of two individuals dated about 500–1100 CE, and harmine in the hair of an infant and an adult from about 400–900 CE. The original hair-study authors read harmine alone as possibly medicinal rather than visionary, since harmine is not itself a classic hallucinogen (Miller et al., 2019). We report the hair studies only as Miller and colleagues summarise them; we did not open the originals. Evidence: preliminary.

Vilca was not only snuffed. At Quilcapampa, a Middle Horizon site in southern Peru (about 600–1000 CE), excavators argue that vilca was consumed alongside a fermented fruit drink at shared events, and that this reinforced the power of the Wari state (Biwer et al., 2022). Evidence: the plant identification is the authors’; the political interpretation is preliminary.

What can residue chemistry prove, and what can’t it?

Mass spectrometry is very good at one thing: confirming that a specific molecule is present in a sample, by matching its mass, fragments and retention time to a reference. The Bolivian team required two fragment transitions per compound and a fixed ratio between them before calling a result positive (Miller et al., 2019). The Chavín team matched peaks against spectral databases, reference standards and published retention data, and ran an independent, blinded microscopy check in parallel (Rick et al., 2025). When two independent methods agree, as at Chavín, the identification is strong. Evidence: established.

What a positive result does not tell you is just as important. Drawing only on the papers above, the main limits are:

• Contact is not use. The Bolivian plant fragment carried cocaine, which its authors attribute to coca leaves rubbing against it inside the pouch, not to the plant itself (Miller et al., 2019).
• One molecule can have several sources. DMT occurs in trace amounts in vilca, so a DMT trace beside bufotenine does not prove a second plant (Miller et al., 2019).
• Absence is ambiguous. At Chavín, tubes with no detectable residue may have held something else, or the residue may simply not have survived (Rick et al., 2025).
• A kit is not a person. Without associated human remains, a residue shows what an object held, not who took it or how often (Miller et al., 2019).
• Harmine is not a drink. A 2024 review concludes that archaeological, botanical, biochemical and analytical considerations make the origin of harmine and harmaline in ancient mummies and artifacts uncertain, and that the evidence does not convincingly prove ancient ayahuasca use (Greco et al., 2024).

That last point is the sharpest live debate. Some authors have treated harmine in Andean finds as a trace of the ayahuasca vine and dated its use to at least 500 CE; the 2024 review argues those readings are not secure (Greco et al., 2024). The Chavín letter adds that tobacco itself contains harmala-type compounds, which further complicates reading harmine as ayahuasca (Whitridge, 2025). For how ayahuasca differs from DMT taken on its own, see ayahuasca vs DMT. Evidence: unresolved.

Common claims about ancient DMT, checked

What we still don’t know

• Whether a drink combining harmine and DMT, like modern ayahuasca, existed before the colonial era (unresolved).
• Where the DMT trace in the Bolivian pouch came from: vilca itself or a separate plant (unresolved).
• Whether the bone tubes found in Chavín households were used for snuff too; their residues have not been tested (unresolved).
• How the plants reached highland sites: trade, travel or local gathering. The Chavín team argues both plants could be reached within about a day’s walk; the commentary points to links with the upper Marañón (preliminary).
• What these substances meant to the people who used them. Chemistry identifies molecules; meaning comes from context and, cautiously, from later ethnography (unresolved).

The fair summary: there is solid chemical evidence that people in the Andes used vilca-based snuffs, containing bufotenine and traces of DMT, for at least two and a half thousand years at Chavín and longer in the south. The leap from that to “ancient ayahuasca” or “ancient DMT trips” is bigger than the evidence. For how the story continues from the Andes into the laboratory, see the history of DMT; for the modern science of the molecule, start with what is DMT; for how a mushroom history was built from similar kinds of evidence, see the history of magic mushrooms.

Shroomverse sells mushroom products. This article is general information about the archaeology of DMT-related plant use. It is not medical, legal or purchasing advice, and it does not recommend any product.