← All articles

Mikhail Polyakov

Could the birch tar of Neanderthals heal wounds? What an experiment with six samples showed and how a scientific paper differs from a retelling in the news

Level: BeginnerReading time: 5 min

Topics: Science, History

A Petri dish with paper discs and glowing amber rings of clear zones beside a piece of birch bark and a lump of dark tar on a stone

Short answer: in a paper in the journal PLOS One dated March 18, 2026, a group of researchers led by Tjaark Siemssen tested whether birch tar suppresses bacterial growth. Two of the methods were reconstructed from finds of the Neanderthal era, and the third is modern. There were six samples in all. Against Staphylococcus aureus, five samples showed a moderate effect, and one did not work at all. Against Escherichia coli, none worked. The authors write that the result is consistent with the medicinal use of tar by some modern peoples, and they call for allowing that in the Neanderthal era tar could have been used not only as a glue. But this is a laboratory experiment and not proof that Neanderthals smeared tar on wounds. The ScienceDaily retelling from October 2026 sounds more confident than the paper itself. Below are the facts and three habits for reading such news.

What was done

All the facts below are taken from the paper by Siemssen and co-authors in PLOS One (21(3): e0343618).

  • What the tar is. Birch tar is known from Neanderthal finds in Germany, on the Dutch coast and in Italy. Its documented use in the Neanderthal era is mainly connected with fastening tools as a glue.
  • How it was made. The bark of two birch species, Betula pendula and Betula pubescens, was turned into tar by three methods: distillation in a tin can (a modern method), distillation in a clay structure, and condensation from burning bark under a stone surface. The last two are reconstructed from Middle Palaeolithic finds.
  • How many samples. Six, BT1-BT6. The tar yield differed greatly: from 0.06 to 0.69 grams.
  • How it was tested. By the paper-disc diffusion method (the modified Kirby-Bauer method): a disc with the substance is placed on a culture of bacteria, and the zone around it where bacteria do not grow is measured. There were two bacteria: Staphylococcus aureus, which is known as a cause of wound infections, and Escherichia coli. The disc had a diameter of 6 mm, so a value of 6 mm means no zone.

What came out

  • Against Staphylococcus. The best sample (BT5, clay-structure method, Betula pendula bark, concentration 133 mg/mL) gave a zone of 10.5 ± 0.7 mm. Four samples (BT1, BT2, BT4, BT6) gave from 7.0 to 7.5 mm. Sample BT3 did not work. In the abstract the authors call the effect moderate.
  • Against Escherichia coli. No sample gave a measurable zone.
  • Controls. The antibiotic gentamicin (10 micrograms per disc) gave zones of 22-23 mm against both bacteria. The solvent (DMSO) without tar did not work.
  • Method of production. The authors write that efficacy is not related to the method of production, and note that all methods show some variation.

How this differs from the retelling

The ScienceDaily retelling, published in October 2026, reports that every type of tar made suppressed the growth of Staphylococcus aureus regardless of the method, and describes two reconstructed ancient methods. The retelling speaks of every type of tar and does not give the number of samples, so from it one cannot tell that one sample out of six (BT3) did not work: this is not necessarily an error of the retelling, but the picture in it is smoother than in the results table. In the paper, a modern tin-can method is added to the two ancient ones, and about Escherichia coli it says there is no effect. In the retelling there are neither the sizes of the zones nor Escherichia coli: only Staphylococcus aureus is mentioned. Siemssen's quote that antimicrobial properties may have played a role already among early Neanderthals is consistent with the authors' conclusion. But the details that change the impression are visible only in the paper.

What the experiment does not show

  • That Neanderthals smeared tar on wounds. The authors suggest considering such use, relying on the similarity with the use of tar by the Mi'kmaq in Canada and by the Saami, but there are no direct traces of treatment in the paper.
  • That the tar is safe or effective for humans. The experiment was run on bacterial cultures of two strains, not on people. The ScienceDaily retelling says directly that suppressing bacteria in a laboratory does not mean that the tar is safe and effective for treating human wounds, and that further research is needed.
  • That the tar acts on different bacteria. It did not act on Escherichia coli.

Three habits when reading news about "ancient medicines"

These are my suggestions, not the authors'.

  1. Open the paper, not the retelling. Who does it: any reader. How to check: using the link or DOI from the retelling, find the journal page and compare the number of samples and the main result.
  2. Look at the number of samples and the result for each. Here there are six different samples and one without an effect; these are six samples, not six repetitions of one experiment. Who does it: the reader. How to check: in the results section find the table and count how many samples gave an effect.
  3. Separate an experiment with bacteria in a dish from a trial on people. Who does it: the reader. How to check: find in the methods section what the experiment was run on (bacteria on a nutrient medium, animals or people); if on bacteria in a dish, a conclusion about treating a person remains a supposition.

It is not advisable to apply these tips to treatment on your own, for example to make tar at home to treat wounds: the paper does not show such benefit for people.

Summary

A laboratory experiment of 2026 showed that birch tar made by three methods (two reconstructed from Neanderthal finds) moderately suppresses the growth of Staphylococcus aureus, but not in all samples, and does not act on Escherichia coli. The authors call for allowing for medicinal use of tar in the Neanderthal era, but there is no direct proof in the paper. The ScienceDaily retelling simplifies the picture.

I work on AI agents and automation. If you would like to see my projects or discuss your task, visit the portfolio.

Sources