Epilepsy is diagnosed through a combination of clinical history, brain wave recording, and imaging, not a single definitive test. The formal definition requires at least two unprovoked seizures more than 24 hours apart, or a single unprovoked seizure with a 60% or greater chance of another one within the next 10 years. Understanding how doctors piece together these different tools helps explain why diagnosis sometimes takes time.
What Counts as an Epilepsy Diagnosis
The International League Against Epilepsy defines epilepsy as a brain disease meeting any of three conditions: two or more unprovoked seizures occurring more than 24 hours apart, one unprovoked seizure with a high probability of recurrence (at least 60% over the next decade), or a recognized epilepsy syndrome. That 60% threshold matters because it means some people can be diagnosed after a single seizure if other evidence, like an abnormal EEG or a brain lesion on imaging, suggests recurrence is likely.
One important detail: multiple seizures that happen within the same 24-hour window are counted as a single event for the purpose of estimating recurrence risk. So if you had three seizures in one day but never again, that alone wouldn’t meet the two-seizure threshold.
The Clinical History
The most important diagnostic tool is a detailed account of what happened before, during, and after the event. Since most people don’t remember their own seizures clearly, doctors often rely heavily on witnesses. They’ll ask about warning signs (unusual smells, visual disturbances, a rising feeling in the stomach), what the body did during the episode, how long it lasted, and how you felt afterward. Confusion, muscle soreness, and memory gaps after an event all point toward epileptic seizures rather than other causes.
Your medical history fills in critical context. A history of head injury, stroke, brain infection, or febrile seizures in childhood all raise the likelihood of epilepsy. Family history matters too, since several genetic forms of epilepsy run in families. Doctors will also ask about sleep deprivation, alcohol use, and medication changes, because these can provoke seizures in someone who doesn’t actually have epilepsy.
Blood Tests to Rule Out Other Causes
Before settling on an epilepsy diagnosis, doctors need to confirm the seizure wasn’t provoked by something temporary. Blood work checks for low blood sugar, abnormal electrolyte levels (sodium, calcium, magnesium), signs of infection, and sometimes drug or alcohol levels. A seizure caused by dangerously low sodium, for example, isn’t epilepsy. It’s a provoked seizure, and treating the underlying problem typically prevents it from happening again.
EEG: Recording Brain Activity
An electroencephalogram records electrical activity in the brain using small sensors placed on the scalp. It’s the single most useful test for confirming epilepsy because it can detect abnormal electrical patterns between seizures, not just during them. These patterns, called epileptiform discharges, include characteristic spikes, sharp waves, and rhythmic bursts that indicate a brain region prone to seizures.
A routine EEG lasts about 20 to 40 minutes and is painless. The limitation is that brain activity is normal between seizures in many people with epilepsy, so a single routine EEG can come back normal even when epilepsy is present. Sleep deprivation before the test or using flashing lights during recording can increase the chance of capturing abnormal activity.
Inpatient Video EEG Monitoring
When a routine EEG isn’t enough, doctors may recommend continuous monitoring in a hospital epilepsy unit. You’re connected to an EEG while video cameras record your behavior around the clock, so doctors can match brain activity to physical symptoms in real time. This is especially valuable when the diagnosis is uncertain or when surgery is being considered.
Diagnostic and classification stays average about 2.3 to 2.4 days, while presurgical evaluations run longer at around 3.5 days. Overall, inpatient monitoring provides a useful diagnosis in roughly 44% of cases, which may sound modest but reflects the fact that many patients referred for monitoring have already been extensively tested without a clear answer.
Brain Imaging With MRI
An MRI scan looks for structural problems in the brain that could be causing seizures: scar tissue, tumors, malformations in brain development, or damage from a prior stroke or injury. A standard brain MRI may miss subtle abnormalities, so epilepsy centers use a dedicated protocol with specific imaging sequences designed to highlight the types of lesions most commonly associated with seizures.
The strength of the MRI magnet makes a real difference. A higher-strength 3-Tesla (3T) scanner detects epilepsy-related lesions about 2.5 times more reliably than the standard 1.5T machines found in many hospitals. In one comparison, 3T imaging identified lesions in 88% of cases versus 74% at 1.5T. If your initial MRI looks normal but seizures continue, a repeat scan on a 3T machine with an epilepsy-specific protocol can sometimes reveal what was missed.
Genetic Testing
Genetic testing has become a standard part of the workup for epilepsy that doesn’t have a clear structural cause. It’s particularly important for seizures that begin in the newborn period, infantile or childhood epilepsy with developmental delays, and frequent febrile seizures. The most commonly identified genetic causes involve genes that control how brain cells communicate electrically, including SCN1A, KCNQ2, and SYNGAP1.
Finding a genetic cause isn’t just academic. Some genetic epilepsies respond better to specific treatments, and certain common medications can actually worsen seizures in specific genetic conditions. A genetic diagnosis also gives families information about inheritance and helps avoid years of unnecessary testing.
Ruling Out Non-Epileptic Events
Not everything that looks like a seizure is one. Psychogenic non-epileptic seizures (PNES) are episodes that resemble epileptic seizures but aren’t caused by abnormal electrical activity in the brain. They’re surprisingly common, accounting for a significant portion of referrals to epilepsy centers, and many people with PNES are initially misdiagnosed with epilepsy and treated with seizure medications that don’t help.
No single feature separates PNES from epileptic seizures with certainty, which is why video EEG monitoring is often needed. However, several patterns tend to differ. Non-epileptic events typically last longer, involve more visible fear or emotional distress during the episode, and are more often followed by crying. Epileptic seizures, by contrast, more often involve self-injury during the event and are followed by amnesia, confusion, and body aches. Other conditions that can mimic seizures include fainting spells (syncope), certain heart rhythm problems, panic attacks, and sleep disorders.
How the Pieces Fit Together
Epilepsy diagnosis is rarely a single-test process. A typical path starts with a detailed history and physical exam, blood work to exclude metabolic or toxic causes, and a routine EEG. If results are inconclusive, the next steps usually include a dedicated epilepsy MRI protocol and possibly extended EEG monitoring. Genetic testing is added when no structural cause is found, especially in younger patients or those with developmental concerns.
The process can feel slow, particularly when routine tests come back normal. But each layer of testing serves a specific purpose: confirming that seizures are truly epileptic, identifying where in the brain they originate, and uncovering an underlying cause that might change treatment. For some people, a confident diagnosis comes after a single office visit. For others, it takes weeks or months of monitoring and imaging to get the full picture.