Traces of ancient "glue" on a stone tool from 50,000 years ago points to complex thinking by Neanderthals, experts say.
The glue was made from birch tar in a process that required forward planning and involved several different steps.
It adds to mounting evidence that we have underestimated the capabilities of our evolutionary cousins.
Only a handful of Neanderthal tools bear signs of adhesive, but experts say the process could have been widespread.
The tool, found in the Netherlands, has spent the last 50,000 years under the North Sea. This may have helped preserve the tar adhesive.
Co-author Marcel Niekus, from the Stichting STONE/Foundation for Stone Age Research in Groningen, said the simple stone flake was probably used either for cutting plant fibres or for scraping animal skins.
While birch tar may have been used by Neanderthals to attach stone tools to wooden handles in some cases, this particular tool probably had a grip made only of tar. Dr Niekus said there was no imprint from a wood or bone shaft in the tar.
It would have enabled the user to apply more pressure to the stone flake without cutting their hands - turning the edge into a precision cutting tool.
The tool was made by Neanderthal groups living at the icy limits of their range, say the authors of the study. At the time, this area would have been part of Doggerland, a landmass that is now subsumed under the North Sea.
These small hunting groups would have inhabited icy tundra, with relatively few trees.
"They had to really plan ahead, because the process needs at least 40kg of wood. In steppe tundra conditions that's not easy to collect, because you only have dwarf birch trees," Dr Niekus told BBC News.
"They also had to invest time and energy in building the fire and extracting the tar."
Researchers used to think Neanderthals only hafted (the action of attaching a handle or strap to a cutting edge) certain types of specialised tools, like points and scrapers.
The Dutch find, along with a few others from Europe, shows that "they also hafted very simple, ugly flakes," said Dr Niekus. "That's something we didn't expect.
"With the investment in time needed, you would expect them only to do it with special hunting weapons, but they did it with special domestic tools as well. We think the use of birch tar was quite widespread."
There are hundreds of Neanderthal sites in the Netherlands, but this is the first Neanderthal birch tar found in the country, and it is hardly ever found in Europe. Marcel Niekus thinks this is because the tar is not preserved under usual conditions. The circumstances under the North Sea were perfect for preserving the tar, providing "a tiny window on Neanderthal normality".
"The important aspect of our find is that we can show that out of the different known methods to distill the pitch from birch bark, Neanderthals used the more complex ones," said co-author Dr Gerrit Dusseldorp from the University of Leiden.
"These are more efficient, and the distribution of contaminants in the tar that we can see on CT-scans is similar to that in complex distillation methods."
Birch tar is also found in Neanderthal contexts at Campitello, Italy, at 200,000 years ago and at Königsaue, Germany, where the evidence is 50,000 years old.
Neanderthals in Italy may also have used pine tree resin for hafting 50,000 years ago. But this natural substance is not as pliable, making researchers think that birch tar was probably their first choice. There are also traces of bitumen found in Neanderthal contexts between 42,000 and 70,000 years ago.
The stone tool was found on Zandmotor beach near The Hague, from the same sandy beds that have yielded a Neanderthal skull fragment. Carbon dating of the tar yielded an age around 50,000 years.
"Modern humans in South Africa are known to produce adhesives from around 100,000 years ago," Dr Dusseldorp told BBC News.
"This is 100,000 years later than the earliest known Neanderthal find. However, because such finds are only rarely preserved this does not definitively prove that there are no older modern human adhesives. We just haven't found them yet."
The paper describing the find is published in Proceedings of the National Academy of Sciences (PNAS) journal.
Follow Paul on Twitter.