Jebel Sahaba: What a Restudy Did to the Oldest Massacre

For fifty years Jebel Sahaba in Sudan was the oldest known massacre. A 2021 re-examination of the same bones found healed wounds alongside fatal ones — and turned one event into a generation of raiding.

6 September 2026 · Part of The Stone Age →

Jebel Sahaba, on the east bank of the Nile in what is now northern Sudan, has been described for over fifty years as the earliest evidence of warfare anywhere. In 2021 a team went back to the same bones, examined all of them under a microscope, and came away with a different story — not a quieter one, but a longer one.

What was found, and when

The cemetery was excavated in the mid-1960s by Fred Wendorf, as part of the salvage archaeology carried out before the Aswan High Dam flooded a stretch of the Nubian Nile. It was published in 1968 as Site 117: a burial ground of the Final Palaeolithic, with stone points found in and among the bodies, and a striking number of individuals showing injuries.

The dating puts the cemetery between roughly 13,400 and 18,600 years ago, which makes it the earliest funerary complex known from the Nile valley. The collection went to the British Museum, and the interpretation went into every survey of prehistoric warfare written since: a massacre, and the oldest one on record.

What the restudy changed

The 2021 re-examination in Scientific Reports did three things.

It raised the count. With supernumerary bones and teeth taken into account, the site can now be regarded as holding the remains of at least sixty-four individuals — three of which are now missing from the collection. The older figure of sixty-one was a floor, not a ceiling.

It found far more injuries. More than a hundred previously undocumented lesions were identified, on individuals already known to be injured and on others who had not been recognised as such. Several had lithic artefacts still embedded in the bone. Most of the trauma is consistent with projectile weapons, which in this period means arrows and darts.

It found healing. This is the finding that matters. A quarter of the skeletons with lesions carry both healed and unhealed trauma — that is, injuries survived long enough for bone to remodel, and injuries that were not survived. A single massacre cannot produce that pattern. People who were shot, recovered, and were later shot again were living through something that lasted.

One event or a generation?

The distinction is not academic. A massacre is an episode: a raid, a rout, a collapse of relations between two groups on a single day. Repeated raiding is a condition — a way of life in which violence recurs at intervals over years, and in which the same individuals can be wounded more than once.

The Late Pleistocene Nile valley offers a plausible setting for the second reading. The river was a corridor of concentrated resources in an increasingly arid landscape, and groups whose subsistence depended on the same fishing and fowling grounds had reason to contest them repeatedly. Embedded stone and healed fractures have also been documented at the Wadi Halfa cemetery associated with the Qadan industry, which suggests the pattern is not confined to one graveyard.

None of that turns Jebel Sahaba into something gentle. It turns it into something more durable, and rather more disturbing: not the memory of a single terrible day, but the ordinary background of several generations.

Why the row says both

Our register of collective violence before 3000 BCE records Jebel Sahaba with a minimum number of individuals that names both figures — the sixty-one in the collection and the at-least sixty-four of the restudy, with the three now missing stated — and with a certainty of medium rather than high.

Medium is the honest rating. The bones are real, the injuries are real, the count is as firm as anyone can make it. What is contested is the shape of the event, and a register that quietly picked one reading would be doing the work a reader ought to be able to do themselves.

The same problem recurs at the other end of the record. Nataruk in Kenya, published in Nature in 2016 as the earliest evidence of inter-group conflict among foragers, was challenged within months by specialists who argued that the cranial damage is consistent with sediment pressure after burial. Our row for Nataruk carries low certainty and says so.

What a restudy is for

There is a temptation to treat a re-examination as a debunking, and this one was reported that way in places. It is better understood as what the discipline is supposed to do: the material stays, the questions change, and the second look asks things the first look could not.

Wendorf's team worked against a rising reservoir with the methods of the 1960s, and produced a record good enough that a team fifty years later could go back to the same collection and find a hundred injuries they had missed. That is not a failure of the original work. It is the reason the original work still matters.

Sources and further reading

The findings described here are from Crevecoeur and colleagues, "New insights on interpersonal violence in the Late Pleistocene based on the Nile valley cemetery of Jebel Sahaba", Scientific Reports 11 (2021), which is open access. Wendorf's original 1968 report remains the description of the excavation itself. For the comparison with a European site where the same single-event question arises, see the 2023 study of San Juan ante Portam Latinam in the same journal. Our dataset of prehistoric collective violence records thirteen sites on the same terms, each with the reference that carries it.

Prefer to listen? Hear this story as a podcast episode or play it on Spotify or Apple Podcasts.

As an Amazon Associate I earn from qualifying purchases. Affiliate links in Recommended reading — and in Further reading on older articles — may earn Writesides of History a commission at no extra cost to you.