Trepanation — deliberately creating an opening in the skull — is one of the oldest surgical procedures known from the archaeological record. Trepanned skulls have been found across Europe, Africa, Asia, the Pacific and the Americas, sometimes with clear evidence that the person survived the operation.
Because epilepsy has a long history of supernatural interpretation, trepanation is often presented in popular accounts as a prehistoric treatment for seizures: a hole was supposedly opened in the skull to allow an evil spirit to escape.
The problem is that archaeology cannot usually tell us that.
A hole in an ancient skull can reveal that an operation was performed. Bone healing can sometimes show that the individual survived. Tool marks may suggest how the opening was made. Associated fractures may support a traumatic indication. But without written evidence, the skull alone usually cannot tell us whether the reason was epilepsy, head injury, headache, infection, ritual practice, post-mortem preparation or something else.
The history of trepanation and epilepsy is therefore most interesting when we separate what is physically demonstrable from what later historians have inferred.
What is trepanation?
Trepanation, also called trephination, is the intentional removal or perforation of part of the cranial vault. Ancient and traditional techniques have included scraping, grooving, cutting and drilling.
The procedure is distinct from accidental skull injury and from holes caused by disease, decomposition or natural anatomical variation. Archaeologists therefore have to distinguish genuine trepanation from what are sometimes called pseudotrepanations — defects that resemble an operation but have another cause.
That distinction is not always easy. Shape, edge morphology, tool marks, location and evidence of healing all contribute to interpretation.
How old is trepanation?
Trepanation is documented in prehistoric remains from the Neolithic period and later, and examples have been reported from widely separated parts of the world.
Large collections have been found in prehistoric Europe and especially in the Andes, while archaeological and ethnographic evidence also comes from Africa, Asia and the Pacific. The broad distribution means trepanation was not a single invention practised in one civilisation and then copied everywhere.
Different societies developed different techniques and probably operated for different reasons.
Some people clearly survived
One of the most striking findings is evidence of healing around some cranial openings.
When living bone survives an operation, its edges can remodel over time. Archaeologists may find smoothing, rounding and new bone formation around the defect. These changes indicate that at least some individuals lived after the procedure long enough for healing to begin.
Modern comparisons with postoperative skull changes have improved understanding of this process, although estimating an exact survival time from archaeological bone remains difficult.
Survival matters because it demonstrates that trepanation was not always performed on corpses. In at least some cases it was a genuine operation carried out on living people.
But survival does not reveal the diagnosis
A healed trepanation tells us that a person lived through cranial surgery. It does not tell us why surgery was attempted.
This is the central limitation of claims linking prehistoric trepanation with epilepsy.
Soft tissues are gone. There is no EEG, clinical history or witness account. Many forms of epilepsy leave no identifiable mark on the skull. A person with recurrent seizures and a person with no neurological disorder at all could leave identical skeletal remains.
For most prehistoric examples, therefore, ‘trepanation for epilepsy’ is a hypothesis rather than a diagnosis established by the archaeological evidence.
Trauma provides a stronger explanation in some populations
Some trepanned skulls also show cranial fractures or other evidence of head injury. In such cases, treatment of trauma becomes a plausible explanation.
This is particularly relevant to Andean material, where researchers have long noted an association between cranial injury and trepanation in some collections.
An opening may have been used to remove damaged bone, relieve pressure, treat complications of a depressed fracture or address persistent symptoms after injury.
Even here, certainty is not automatic. A fracture and a trepanation on the same skull do not always establish which came first or exactly what the operator intended. But physical trauma provides a more direct archaeological clue than epilepsy does.
Where did the ‘evil spirits escaping through the hole’ story come from?
The explanation is repeated so often that it can sound like established archaeological fact.
In reality, it is largely an interpretation developed by later scholars trying to explain prehistoric surgery.
In the nineteenth century, the French physician, anthropologist and neuroscientist Paul Broca became deeply interested in prehistoric trepanation. After examining ancient skulls, he proposed that some operations may have been performed for convulsions and that prehistoric people might have believed these attacks were caused by spirits that could be released through the skull.
Broca’s theory was ingenious, but it was still a theory created thousands of years after the operations themselves.
It should not be presented as though a Neolithic surgeon left a written note saying, ‘this hole was made to release a demon causing epilepsy.’
Paul Broca and the epilepsy hypothesis
Broca’s work in the 1860s and 1870s played a major role in establishing that prehistoric trepanation had sometimes been performed on living people.
He also tried to explain why it might have begun.
One hypothesis was that convulsions in children may have been interpreted as evidence of supernatural intrusion. If the child later recovered naturally after trepanation, the operation could appear successful and the practice could spread.
This idea strongly influenced later histories of neurosurgery and epilepsy.
But modern historical writing needs to keep two claims separate: Broca helped establish that ancient trepanation was real surgery; his proposed reason for prehistoric people doing it cannot be independently proved in most cases.
Archaeology and historical texts provide different kinds of evidence
For prehistory, the evidence is material: skulls, tools, burial context and healing.
Once written medical traditions appear, the situation changes. Texts can tell us what particular authors thought cranial surgery was for.
This is why the earliest reasonably secure links between epilepsy and trepanation come from written antiquity rather than from prehistoric skulls alone.
The Hippocratic tradition
Ancient Greek medical writing contains discussion of opening the skull, especially in relation to head injury. Historical reviews of epilepsy surgery also point to the Hippocratic tradition as an early written source connecting cranial intervention with seizure disorders.
The underlying physiology was humoral rather than neurological in the modern sense. Disease might be explained through phlegm, bodily fluids and obstruction.
This is different from the modern concept of removing or disconnecting an epileptogenic brain region. Trepanation opened the skull; it was not yet modern epilepsy surgery.
The distinction matters because the history of epilepsy surgery should not be pushed backwards simply because both practices involve the head.
Aretaeus and later classical medicine
Later Greco-Roman medical writers continued to discuss trepanation among possible treatments for severe disorders, including epilepsy in some historical accounts.
Aretaeus of Cappadocia is frequently cited in histories of epilepsy as recommending trepanation alongside dietary and medicinal measures.
Such references provide firmer evidence that epilepsy could become an explicit indication for cranial intervention in written medicine.
They still tell us nothing certain about the motivation behind a particular prehistoric skull from centuries or millennia earlier.
Different techniques in different places
Trepanation was not one standard operation.
Scraping could gradually thin the skull until an opening formed. Grooving involved cutting around a defined area until a piece could be removed. Drilling could create one or several perforations. Other techniques used straight cuts.
The location, size and number of openings also varied. Some individuals had more than one trepanation.
This technical diversity is another reason to avoid one universal explanation. A practice used for warfare-related head injury in one society might have had a different role elsewhere.
The Andes and high survival rates
Peru and the wider Andean region have produced some of the richest archaeological evidence for ancient trepanation.
Studies of Andean skulls have documented changing techniques and substantial evidence of survival in many cases. Some periods and communities appear to have developed considerable practical skill in avoiding lethal injury to the skull and underlying tissues.
Head trauma is an important part of the interpretation of these remains, particularly in populations where cranial injury was common.
It would therefore be misleading to describe Andean trepanation primarily as an epilepsy treatment simply because epilepsy appears in general histories of trepanation.
Africa and the survival of traditional trepanation
Ethnographic and historical reports from parts of North and East Africa show that forms of trepanation persisted into much more recent times.
These records are valuable because living practitioners and patients can sometimes explain indications directly.
A 2024 review of prehistoric and traditional cranial operations notes that indications varied greatly by region. In some African settings, persistent headache after trauma was a common reason. In parts of Melanesia, fracture was central. In New Ireland, reports included epilepsy and some forms of mental disturbance.
This variation is crucial: even where epilepsy was genuinely an indication in one documented society, it does not follow that every trepanned skull elsewhere had the same purpose.
Epilepsy in ethnographic accounts
More recent traditional societies give historians something prehistoric archaeology cannot: reported reasons for the operation.
Where epilepsy is explicitly named, the association between seizures and trepanation becomes evidence rather than speculation.
But ethnographic analogy has limits. Historians cannot assume that a twentieth-century practice in one population explains a Neolithic operation in another continent thousands of years earlier.
Similarity can suggest possibilities; it cannot prove identity of motive.
Could some prehistoric patients have had seizures after head injury?
Yes, and this creates an interesting overlap between two proposed indications.
A person who suffered a severe cranial injury could later develop recurrent seizures. A community might then trepan the skull because of the original injury, the later attacks, persistent headache, neurological symptoms or some combination of them.
From the skeleton alone, it may be impossible to reconstruct that sequence.
This is one reason modern diagnostic labels should be used carefully in paleopathology.
Multiple trepanations
Some skulls contain more than one opening, and some openings show different stages of healing.
This demonstrates that certain individuals underwent repeated cranial procedures.
Repeated operations have sometimes been interpreted as evidence of chronic conditions such as epilepsy or recurrent headache. That is possible, but repetition alone does not establish the reason.
A person could also experience multiple injuries, infection, repeated ritual procedures or staged surgical treatment.
Post-mortem trepanation
Not every cranial opening was necessarily therapeutic.
Researchers have considered post-mortem trepanation, removal of skull pieces as amulets, ritual preparation and other mortuary practices among possible explanations for some specimens.
Distinguishing an operation performed shortly before death from one performed immediately after death can be difficult when there is no time for healing.
This is another reason archaeological interpretation must be cautious.
The problem of pseudotrepanation
Not every hole in an ancient skull was deliberately made by a surgeon.
Congenital defects, infection, tumours, erosion after burial, animal activity and other processes can create openings or irregularities.
Modern bioarchaeological work therefore examines the geometry and margins of suspected trepanations carefully.
Research comparing archaeological specimens with modern postsurgical bone healing has shown how complicated this diagnosis can be.
A dramatic-looking skull defect is not automatically evidence of prehistoric neurosurgery.
Can we estimate whether someone survived?
Broadly, yes: substantial bone remodelling indicates survival beyond the immediate operation.
Precisely, it is harder.
The rate of cranial healing varies with age, health, wound size, infection and other factors. Archaeological preservation can also obscure the original appearance.
Modern imaging studies have helped build more realistic models of postoperative bone change, but claims such as ‘this person survived exactly six months’ are usually much less secure than the simpler conclusion that healing occurred.
What trepanation does prove
It proves something remarkable even without an epilepsy diagnosis.
Ancient communities developed practical methods for deliberately opening the skull. Some patients survived. Operators learned where and how to cut or scrape while avoiding immediately fatal damage.
That represents substantial empirical surgical knowledge.
The achievement does not need the claim that prehistoric people understood epilepsy as a brain disorder in order to be historically significant.
What trepanation does not prove
It does not prove that prehistoric societies knew the electrical basis of seizures.
It does not prove that every operation was performed to treat epilepsy.
It does not prove that every operator believed evil spirits were trapped inside the skull.
It does not prove that a healed operation cured whatever condition led to the procedure.
And it does not make prehistoric trepanation equivalent to modern epilepsy surgery.
Why the epilepsy connection became so attractive
Epilepsy has historically been associated with possession, supernatural force and dramatic changes in consciousness. Trepanation creates an obvious opening through which a spirit could be imagined to leave.
That makes the explanation narratively satisfying.
It also connects neatly with nineteenth-century ideas about ‘primitive’ religion and with later histories that wanted a simple origin story for neurosurgery.
But a compelling story is not the same as direct evidence.
From trepanation to epilepsy surgery
The true transition to modern epilepsy surgery occurred much later, when physicians began to connect specific seizure patterns with specific regions of the cerebral cortex and surgeons attempted operations directed at the source of seizures rather than merely opening the skull.
That nineteenth-century development belongs to a separate chapter in this history.
Trepanation is its distant ancestor only in the broad sense that both involve cranial surgery. The intellectual basis, targets and goals were fundamentally different.
The central historical lesson
Trepanation gives us extraordinary physical evidence of ancient surgery, but it also demonstrates the limits of archaeological interpretation.
We can see the hole. We can sometimes see the tool marks. We can sometimes see healing.
What we usually cannot see is the diagnosis written in the mind of the person who made it.
Epilepsy may have been one reason for trepanation in some societies, and later written and ethnographic sources show that it certainly became an indication in particular contexts. But prehistoric trepanned skulls alone cannot tell us that seizures were the reason.
That distinction makes the history more accurate — and more interesting — than the familiar story of simply drilling holes to let demons escape.
Sources and further reading
Schijns OEMG et al. The start and development of epilepsy surgery in Europe: a historical review. Acta Neurochirurgica. 2015.
Nerlich AG et al. Cranial trepanation and healing process in modern patients — bioarchaeological and anthropological implications. 2020.
Ganz JC. Prehistoric or current primitive cranial operations. Progress in Brain Research. 2024.
Clower WT, Finger S. Discovering trepanation: the contribution of Paul Broca. Neurosurgery. 2001.
Faria MA. Neolithic trepanation decoded — a unifying hypothesis. Surgical Neurology International. 2015.
The mystery of prehistoric trepanations: is neurosurgery the world’s eldest profession? Neurochirurgie. 2010.
3D reappraisal of trepanations at St. Cosme priory between the 12th and 15th centuries, France. International Journal of Paleopathology. 2021.
Historical Sources on Epilepsy Surgery: From Antiquity Through the End of the 20th Century. The Open Neurology Journal. 2021.
Historical evidence checked: September 2026.