In Morocco, scientists rebuilt a 315,000-year-old Homo sapiens face that looks surprisingly familiar

A team of researchers has built a digital face for some of the oldest known Homo sapiens fossils, and the result looks far more like a person you could meet today than a distant ancestor. The fossils come from Jebel Irhoud in Morocco and date to about 315,000 years ago. The new facial approximation appeared in the journal Heritage on 13 September 2026, and ScienceAlert brought it to a wider audience in October.

The fossils that moved our origin story back

The story of Jebel Irhoud starts in the early 1960s, when a worker extracting minerals at the site found a skull. Researchers first took it for an African Neanderthal about 40,000 years old. Two studies in 2017 changed that picture. They identified the remains as early Homo sapiens and pushed back the known history of our species by about 100,000 years. The date of the fossil layer carries an uncertainty of about 34,000 years, so 315,000 is a best estimate and not an exact figure.

Earlier analyses of these fossils had already noticed something odd. The face looked modern, but the braincase looked archaic. That mix makes Jebel Irhoud a useful window into how our species took shape, and it makes the question of what these people looked like harder to answer than it seems.

A skull assembled from more than one person

The face does not belong to a single individual. The Max Planck Institute for Evolutionary Anthropology in Leipzig combined several fossil specimens into a composite digital skull from the site’s Layer 7. The skull known as Irhoud 1 supplies most of the braincase and the upper face. The lower jaw and teeth come from other individuals found in the same layer. The authors say plainly that the composite does not represent Irhoud 1 alone, and they note that no detailed report documents how the pieces fit together.

How the face took shape

Cícero Moraes, a Brazilian 3D designer, led the work with colleagues from Flinders University in Australia, the University of Pavia in Italy and other institutions. The team used free, open-source tools. They started from a 3D model of the composite skull that the Max Planck researchers built from video frames, then checked its scale against published measurements. They worked in Blender with the OrtogOnBlender add-on, a method Moraes has developed and tested since at least 2013.

Because no ancient skin or muscle survives, the team needed a stand-in. They took a CT scan of an anonymous modern adult and deformed it to fit the Irhoud skull. They set soft-tissue thickness using reference data for people of African ancestry. Then they added a nose shape, adjusted the midface, sculpted a neutral expression and produced two versions: a grayscale image that focuses on the bones, and a colored version.

As a sanity check, the team measured the inside of the digital skull. The brain cavity came out at 1,364 milliliters, close to an earlier CT-based estimate of 1,375 milliliters and well within modern human variation. That match supports the skull model’s proportions. It does not prove the face is accurate.

What this face can and cannot tell us

The authors list their limits openly. They write that no empirical data exist on soft-tissue thickness in Middle Pleistocene humans, so modern people serve as a proxy. That proxy is a necessary shortcut, not proof that ancient and modern anatomy match. The jaw and teeth come from different individuals, and the team left a dental mismatch unadjusted. No direct way exists to validate the result, and the method produces no quantitative error measure.

The colored version carries extra guesswork. Skin tone, hair, facial hair and lighting come from artistic choices, because fossils cannot reveal them. Moraes himself calls the colored image the most artistic version, with fewer scientifically robust features than the grayscale one.

Why it still matters

Even with those caveats, the reconstruction gives a concrete, careful answer to a question people ask about every ancient fossil: what did this person look like? Here the answer is a face that reads as modern, set on a skull with older features. That fits what researchers saw in the bones years ago. It also reminds us that the line between “early” and “modern” humans was blurry, and that our species did not appear overnight with every modern trait in place.

Sources: Moraes, C., Varotto, E., Habicht, M. E., Galassi, F. M., Moura, L. and Abdullah, J. Y. (2026), “Digital Facial Approximation of Jebel Irhoud Composite Skull (~315,000 BP),” Heritage 9(9), 369 (DOI: 10.3390/heritage9090369), and ScienceAlert, 21 October 2026. Lead image: digital facial approximation by Cícero Moraes.