Myths & Claims
Claim: “My EEG was normal, so it cannot be epilepsy.”
Verdict: false. A routine EEG can support an epilepsy diagnosis, but a normal recording does not rule epilepsy out.
Why the myth sounds believable
An EEG records electrical activity from the scalp during the period of the test. If no seizure happens and no interictal epileptiform activity is captured, the recording may look normal even in someone who genuinely has epilepsy.
Epilepsy Society states explicitly that a normal EEG does not exclude epilepsy and that no single test can confirm or rule it out by itself.
What clinicians use instead
Diagnosis combines the event history, witness descriptions or video when available, neurological assessment, EEG findings, imaging and other tests where relevant. Sometimes sleep-deprived, ambulatory or prolonged video-EEG is useful when the first routine EEG is unrevealing.
The reverse mistake matters too
An abnormal EEG does not automatically prove that every unusual event is an epileptic seizure. EEG findings must be interpreted in clinical context.
A normal EEG is a normal recording — not proof that epilepsy is absent
An EEG samples electrical activity during the period in which the recording is running. Most routine outpatient EEGs are relatively short compared with the hours, days or weeks between a person’s seizures.
Epileptiform discharges are intermittent. A person can therefore have epilepsy and produce no epileptiform abnormality during one routine recording. NICE explicitly advises that EEG should not be used to exclude a diagnosis of epilepsy.
Why can epilepsy be invisible between seizures?
Many EEGs are recorded interictally — between seizures. The purpose is often to look for interictal epileptiform discharges such as spikes, sharp waves or spike-and-wave complexes. These can support a diagnosis and help classify the epilepsy, but they are not continuously present in everyone with epilepsy.
Some people generate frequent interictal abnormalities. Others produce them mainly during sleep, shortly after waking, after sleep deprivation or only occasionally. Some focal epilepsies generate activity in brain regions that scalp electrodes detect poorly.
How sensitive is a first routine EEG?
The exact yield depends on the population, epilepsy type, timing, recording length and what counts as an abnormal result. Across older studies of people with established epilepsy, epileptiform abnormalities were found on roughly 29–55% of first routine EEGs. In population-based work, the yield increased substantially when EEGs were repeated.
After a first unprovoked seizure, sensitivity can be even lower. A 2016 systematic review and meta-analysis found that epileptiform discharges on a routine EEG had high specificity in adults but low sensitivity. That is exactly why a negative test cannot safely rule epilepsy out.
Why timing matters
EEG yield can be higher when the recording is performed relatively soon after a seizure. NICE recommends that, when an EEG is requested after a first suspected seizure, it should be performed as soon as possible — ideally within 72 hours.
This does not mean an EEG performed later is useless. It means the chance of capturing useful interictal abnormalities may change with timing.
Why sleep can reveal what wakefulness misses
Sleep changes the way large neuronal networks synchronise. In several epilepsy syndromes and focal epilepsies, interictal epileptiform discharges become more frequent or more visible during drowsiness and sleep.
That is why a sleep EEG or sleep-deprived EEG may be considered when a routine awake EEG is unrevealing. NICE recommends considering a sleep EEG after a normal routine EEG when diagnostic uncertainty remains.
What does sleep deprivation actually do?
Sleep deprivation can help in two ways. It increases the likelihood that the person will sleep during the recording, and sleep deprivation itself may increase the occurrence of interictal epileptiform discharges in susceptible people.
It is not simply a longer routine EEG. It deliberately changes the physiological state being sampled.
Why longer recordings can help
A 20–40 minute routine EEG examines only a small fraction of a person’s day. Extending the recording increases the opportunity to capture drowsiness, sleep and intermittent epileptiform activity.
Studies comparing routine with extended recordings show that some epileptiform discharges first appear after the initial routine recording window. The diagnostic gain is not unlimited, but longer sampling can matter when abnormalities are infrequent.
Why repeating EEG can change the answer
One normal EEG is one sample. A second or third recording samples different periods of brain activity and may include different sleep–wake states.
Reviews of serial EEG studies show a substantial rise in the cumulative detection of interictal epileptiform discharges after repeated recordings. That does not mean everyone with a normal EEG needs endless repeat tests. Repetition is most useful when the diagnosis or classification remains uncertain and another EEG could genuinely alter management.
Ambulatory EEG: more time in ordinary life
Ambulatory EEG allows recording over many hours — often a day or longer — while the person is outside a conventional EEG laboratory. This increases the chance of capturing sleep, interictal discharges or one of the person’s habitual events.
NICE recommends considering ambulatory EEG, for up to 48 hours, when routine and sleep EEG are normal but diagnostic uncertainty persists.
In a 2023 Class I diagnostic study of adults after a first single unprovoked seizure, 24-hour ambulatory EEG detected interictal epileptiform discharges or seizures more often than either the first or second routine EEG. That study concerns a specific clinical population, so its exact percentages should not be applied to everyone with suspected epilepsy, but it demonstrates why longer sampling can increase sensitivity.
Video-EEG answers a different question
Long-term video-EEG combines continuous EEG with synchronised video. Its greatest value is often not simply finding an interictal spike. It can record one of the person’s actual attacks and compare the visible behaviour with the electrical activity at the same moment.
This can be particularly important when clinicians are trying to distinguish epileptic seizures from functional/dissociative seizures, syncope, sleep disorders or other episodic events.
But even scalp video-EEG has limits
A seizure can sometimes be clinically real without producing an obvious scalp EEG change.
This is especially relevant for some brief focal aware seizures and seizures arising from deep, mesial or poorly oriented cortical regions. Scalp electrodes record voltage differences that have propagated through brain tissue, cerebrospinal fluid, skull and scalp. A small or deep source may not generate a sufficiently large or suitably oriented field at the scalp.
For this reason, specialists interpret the EEG together with seizure semiology, imaging and the rest of the clinical evidence rather than applying a rule that ‘no scalp change means no seizure’.
Deep focal seizures can be particularly difficult
Temporal, insular and other deep focal epilepsies may present a special challenge. Electrical activity can begin in tissue relatively far from scalp electrodes and only become clearly visible after the discharge spreads.
This is one reason a normal routine EEG does not contradict a convincing focal seizure history. In selected drug-resistant cases where surgery is being considered, specialists may eventually use intracranial methods such as SEEG to investigate networks that scalp EEG cannot localise adequately.
What activation procedures are trying to do
Routine EEG may include hyperventilation and intermittent photic stimulation. These are not attempts to provoke a convulsive seizure for its own sake. They are standard activation procedures designed to increase the chance of revealing characteristic EEG abnormalities in susceptible epilepsy types.
Hyperventilation is particularly useful in some generalised epilepsies, including typical absence epilepsy. Photic stimulation can reveal a photoparoxysmal response in people with photosensitivity. Neither procedure detects every epilepsy.
Medication can influence what is recorded
Antiseizure medicines can reduce seizures and may influence epileptiform activity, although the relationship is not simple enough to use routine EEG as a general ‘medication level’ test.
People should not stop or alter antiseizure medication simply to try to make an EEG abnormal unless this is being done under a specialist-controlled monitoring protocol.
A normal EEG after treatment does not prove epilepsy has gone
Someone with established epilepsy may later have a normal EEG. This can happen because interictal abnormalities are intermittent, because treatment suppresses epileptic activity, or because the recording simply did not sample an abnormal period.
Whether epilepsy is in remission, whether medication can be withdrawn and what the recurrence risk may be are separate clinical questions. They are not decided by one normal EEG alone.
The opposite error: an abnormal EEG does not automatically equal epilepsy
EEG interpretation has two diagnostic traps, not one.
Some people without epilepsy have benign normal variants or uncommon epileptiform-appearing abnormalities. Artefacts from muscle, eye movement, electrodes or movement can also resemble cerebral waveforms. Over-reading EEG is a recognised cause of epilepsy misdiagnosis.
That is why an EEG finding must make sense in the clinical context. A person should not normally acquire an epilepsy diagnosis solely because one waveform was labelled ‘sharp’ on a report.
Epileptiform discharges are supportive evidence
When clear interictal epileptiform discharges are present in the right clinical context, they can strongly support an epilepsy diagnosis and help classify the seizure type or syndrome.
EEG therefore has an asymmetry: a convincing positive finding can be highly informative, while a negative routine recording is much less powerful for excluding epilepsy.
EEG does not ‘see epilepsy’ directly
Epilepsy is diagnosed from the person’s tendency to have epileptic seizures and, in some situations, from the recurrence risk associated with an epilepsy syndrome or enduring brain condition. EEG is one source of evidence within that assessment.
It can show interictal epileptiform discharges. It can sometimes capture an electrographic seizure. It can help distinguish focal from generalised epilepsies and identify particular electroclinical syndromes.
But the diagnosis still begins with the clinical event: what happened before, during and after it, witness accounts or video where available, examination and appropriate tests for alternative causes.
Why a normal EEG can actually be reassuring in one sense
A normal routine EEG means that no relevant abnormality was identified during that particular recording. That is real information. It should not be dismissed as meaningless.
What it cannot do is carry more diagnostic weight than the test possesses. ‘Normal during this sample’ and ‘this person cannot have epilepsy’ are two completely different conclusions.
What happens when the EEG is normal but the history still strongly suggests epilepsy?
The next step depends on the individual case. A specialist may review the original history and witness description, examine any phone video, reconsider mimics, review MRI or other investigations, repeat the EEG, obtain sleep during EEG, use ambulatory monitoring or arrange prolonged video-EEG.
Not everyone needs all of these tests. The purpose is not to keep testing until something becomes abnormal. It is to obtain the evidence needed to answer the remaining clinical question.
What if several EEGs are normal?
Repeated normal studies reduce the chance of finding interictal epileptiform discharges, but they still do not create an absolute exclusion rule.
At that point, specialists may ask a broader question: does the overall event description still make epilepsy the most coherent diagnosis, or has the accumulating evidence shifted the balance toward another explanation?
That is very different from saying ‘the EEG was normal, therefore the symptoms were not real’.
Related pages on The Untold Side
EEG Explained Properly
Diagnosis and Tests
Functional/Dissociative Seizures and Epilepsy
SEEG: Recording Seizures From Inside the Brain
History of EEG and Epilepsy
Sources
NICE NG217 — Epilepsies in children, young people and adults
NICE Evidence review 3 — Diagnosis of epilepsies
Tatum WO. When should we obtain a routine EEG while managing people with epilepsy? 2021.
Bouma HK et al. The diagnostic accuracy of routine electroencephalography after a first unprovoked seizure. 2016.
Hernandez-Ronquillo L et al. Diagnostic Accuracy of Ambulatory EEG vs Routine EEG in Patients With First Single Unprovoked Seizure. 2023.
Epilepsy Diagnosis When the Routine Ancillary Tests Are Normal. Review.
ILAE — Epilepsy diagnosis when the routine ancillary tests are normal
Epilepsy Society — Diagnosing epilepsy
EEG Explained Properly