Scientists have long known that modern humans (Homo sapiens) shared a common ancestor with Neanderthals (Homo neanderthalensis, our closest extinct relative), but no one knew much about what that ancestor looked like…until now. In a new study, researchers from Cambridge University used advanced digital techniques to create a “virtual fossil,” a 3D model of the skull of the last common human-Neanderthal ancestor. In addition to giving scientists a glimpse at the cranium of this mysterious prehistoric predecessor, the study revealed that the lineages appear to have split some 300,000 years earlier than previously thought.

Though it is widely accepted that Homo sapiens and Homo neanderthalensis emerged from a common ancestor sometime during the Middle Pleistocene era—anywhere between 800,000 and 100,000 years ago—we know little about what this early human ancestor may have looked like. Fossils from the period when the split occurred are extremely rare, and those that do exist are fragmentary. To help fill this gap, a team of researchers at Cambridge University decided to apply an innovative digital technology known as “morphometrics,” along with statistical algorithms, to an array of cranial fossils spanning nearly 2 million years in order to virtually recreate the skull of the last ancestor shared by modern humans and Neanderthals.

To create this “virtual fossil,” the scientists plotted a total of 797 so-called “landmarks” on 15 fossilized skulls, including a 1.6-million-year-old Homo erectus fossil, Neanderthal crania found in Europe and 19th-century skulls from the Cambridge collections. These fossils dated to both before (H. erectus was the first species of human to migrate out of Africa, around 1.9 million years ago) and after the human-Neanderthal split. After using the data to create a morphological timeline of how our ancestors developed, the researchers digitally scanned a modern skull and fed it into the timeline, reshaping it to match the different landmarks going back through time.

The team generated three possible ancestral skull shapes, each corresponding to a different predicted time when the human-Neanderthal lineages split. Next, they digitally rendered the three skulls and compared them to the few original fossils and bone fragments that exist from the time period, which allowed them to find the most likely timeframe for the last common ancestor to have roamed Earth. Though previous estimates based on DNA evidence had predicted that the evolutionary split occurred some 400,000 years ago, the new study settled on an earlier time window, around 700,000 years ago.

Though the population of the last common human-Neanderthal ancestor was present across Europe and Eurasia, the researchers say it most likely originated in Africa. They believe the ancestor belonged to the early human species Homo heidelbergensis in its broadest sense. This early Homo species, known as Heidelberg man, lived in Africa, Europe and western Asia between 700,000 and 300,000 years ago.

The new study’s findings, published last week in the Journal of Human Evolution, suggests that our long-lost ancestor had the heavy, thickset brow characteristic of early Homo species like Neanderthals but lost in modern humans. The digitally reconstructed skull also showed the beginnings of a bulge in the back, evidence of what would become the “occipital bun” that contributed to the elongated shape of the Neanderthal head. On the other hand, the early human ancestor’s face shows indentations under the cheekbones like modern humans have, differentiating them from earlier hominid species. In Neanderthals, the bones in this area (the maxillia) were thicker due to air pockets underneath, causing the face to protrude.

The scientists now plan to apply their digital reconstruction methods to build a model of the last common ancestor of Homo and chimpanzees, and they believe the technology has even broader scientific potential. According to lead author Dr. Aurélien Mounier: “Our models are not the exact truth, but in the absence of fossils these new methods can be used to test hypotheses for any paleontological question, whether it is horses or dinosaurs.”