Why a painting is hard to date at all
A painting on a cave wall carries no label. Most of prehistoric art was made with mineral pigments, red and yellow ochre or black manganese, and minerals don’t record the moment someone spread them on rock. What you can date is something organic in the paint, or something that formed afterwards on top of it.
That is why every date for a cave painting comes from an indirect clue. The skill of the dating team lies in choosing the clue, and in being honest about what it can and can’t prove.
Radiocarbon: dating the charcoal itself
Some black lines were drawn with charcoal and manganese, and charcoal is burnt wood, so it contains carbon from a once-living tree. Radiocarbon dating measures how much of the unstable form of carbon, carbon-14, is left in it. Because carbon-14 decays at a known rate, the amount left tells you when the tree died.
The method has three limits. It works only on organic pigments, so a red ochre animal can’t be dated this way. It destroys the sample, so teams take tiny scrapings, and with tiny samples a little contamination makes a big difference. And it stops at about 40,000-50,000 years, when too little carbon-14 is left to measure.
Even when it works, there is a catch. When the Chauvet team published 88 new radiocarbon dates in 2016, the archaeologist Paul Pettitt replied that dating charcoal proves only the age of the charcoal. A painter could have picked up a piece of old burnt wood, or drawn on a wall long after a fire. For this reason the strongest dates combine several clues.
Uranium-thorium: dating the crust on top
So how do you date a red animal with no carbon in it? You look at what grew over it. Water seeping through limestone caves leaves behind thin layers of calcite, the same mineral that builds stalactites, and sometimes those layers form a crust over the paint. That water carries traces of uranium. Once the uranium is locked inside the calcite, it slowly decays into thorium, and the ratio between the two works like a clock that starts when the crust forms.
The method reaches much further back than radiocarbon, to about 500,000 years. In 2012 a team led by Alistair Pike used it on eleven caves in northern Spain and found a red disk at El Castillo with a minimum age of 40,800 years. At Altamira the same work showed that painting went on, on and off, for up to 20,000 years. In 2024 a laser-based version of the method dated the pig scene at Leang Karampuang, on the Indonesian island of Sulawesi, to at least 51,200 years ago.
Why “at least” matters
A crust can only form after the painting it covers. So its date is a floor: the painting is at least that old. It could have been painted the day before the crust began to grow, or thousands of years earlier. Put simply, uranium-thorium tells you the latest possible date for a painting, never the earliest.
Only in rare cases can a team box a painting in from both sides, when there is a dated layer underneath the paint as well as one on top. Most of the time you get a minimum, and that is why records for “the oldest art” keep getting older as more crusts are sampled.
When the methods disagree
Uranium-thorium has its critics. It assumes that no uranium was washed in or out of the crust after it formed; if water dissolved part of the calcite, the clock is off. At Nerja, in southern Spain, a crust over a charcoal mark gave an age of about 119,000 years, while radiocarbon on the charcoal itself gave about 19,000. Some researchers argue that crusts which dissolved and re-formed tend to look too old.
The quarrel has real consequences. The claim that Neanderthal painters worked in Spanish caves more than 64,800 years ago rests on uranium-thorium dates, and the critics of that claim question exactly these assumptions.
Style, layers and footprints
Before laboratory dates existed, cave art was dated by style: a horse drawn one way was thought to be older than a horse drawn another way. Style still matters, and some specialists use it, together with animal bones and the layers of the cave floor, to argue for younger dates at Chauvet than the radiocarbon suggests.
Other clues come from what people left behind as they worked: hearths, lamps, torch marks on walls, even footprints in clay. None of these dates a painting directly, but they show when people visited the cave. On the ceiling of Altamira and on the walls of Chauvet, the best dates come from putting all of these together: no single number settles the age of a painting.
Where to see it
The paintings dated in these studies are mostly in caves that are closed or strictly limited. You can see the Chauvet drawings in the full-scale replica, Grotte Chauvet 2, in the Ardèche, and the Altamira ceiling in the replica, the Neocueva, at the museum near Santillana del Mar.
Sources
- Pike, A. W. G. et al., “U-Series Dating of Paleolithic Art in 11 Caves in Spain”, Science 336 (2012), 1409-1413.
- Oktaviana, A. A. et al., “Narrative cave art in Indonesia by 51,200 years ago”, Nature (2024).
- Quiles, A. et al., PNAS (2016); Pettitt, P. and Bahn, P., “An alternative chronology for the art of Chauvet cave”, Antiquity (2015).
- Royal Society of Chemistry, “Defining the age of a rock or cave painting”; Current Archaeology, “Science Notes: Palaeolithic cave art and uranium-thorium dating”.
- Bednarik, R. G., Rock Art Research (2022); Archaeology Magazine, “Scientist questions dating of world’s oldest cave art” (2026).
