Can You Stop Epilepsy Medicine After Being Seizure-Free?
27 Oct 2026
Some people can come off epilepsy medicine, but only with their neurologist and only after a long seizure-free stretch. A new meta-analysis of 22 studies found a higher chance of relapse if you stop after less than five years without seizures, and children and adults seem to differ [1]. Never stop or cut down tablets on your own.
Can You Stop Epilepsy Medicine After Being Seizure-Free at a glance
- New evidence: a 2026 systematic review pooled 22 studies and 3,933 people who stopped anti-seizure medicine [1].
- Shorter seizure-free spells carried more risk: under 5 years overall, under 3 years in children, under 5 years in adults [1].
- Many children do well in the long run: 77% of one trial's follow-up group were seizure-free off treatment decades later [2].
- Tapering speed: one small trial found no clear difference between a 60-day and a 160-day taper, but it was too small to settle the question [3].
- The decision belongs to your treating neurologist, not to the internet and not to a clinic that adds supportive care.
What did the new study find?
A team from China pooled published studies on people who stopped anti-seizure medicine (ASM) after a period without seizures. They searched three databases up to November 2025. Of 45 studies identified, 22 covering 3,933 people could be combined in the meta-analysis [1].
The main result: people with less than five years of seizure freedom had a higher risk of relapse than those who had waited longer (relative risk 1.49, 95% CI 1.17 to 1.88) [1]. Put simply, stopping early meant seizures came back about 1.5 times as often as when people waited longer.
How long should you be seizure-free before stopping?
There's no single number that fits everyone. The review split the data by age. Among children, the risk was higher when they had been seizure-free for under three years (RR 1.70, 95% CI 1.20 to 2.39). Among adults, it was higher under five years (RR 1.52, 95% CI 1.02 to 2.26) [1].
The authors also reported results by region and a signal that very long seizure-free periods, ten years or more, were safer. They themselves say that last finding needs checking because few people and few studies were behind it [1]. So treat it as a hint, not a rule.
Does it matter how long a child stayed on medicine?
Possibly less than parents fear. In an older trial, children with epilepsy of unknown cause were randomised to one year or three years of medicine before withdrawal. Researchers traced them 29 to 35 years later. Of the 149 who answered, 110 (77%) were seizure-free without treatment, and the length of the first course made no difference to later epilepsy risk [2].
It's reassuring. But the children had no major disability and the follow-up was a survey, so don't assume it holds for every child. The same study found that adults who still had active epilepsy were more often unemployed and reported poorer mental health [2]. That's one reason doctors take both relapse and quality of life seriously.
Does a slow taper beat a fast one?
Most families assume a slower taper must be safer. The evidence for that is thin. The RASLOW trial randomised 48 adults, all seizure-free for more than two years, to a 60-day or a 160-day withdrawal of single-drug treatment. Within a year, 3 of 25 in the slow group and 1 of 23 in the rapid group had a seizure. Nobody had status epilepticus, a seizure-related injury or died [3].
With numbers this small, the trial can't tell you which approach is better. Your neurologist will choose the schedule from your seizure type, EEG, scan findings and the drug you take.
What should you do before asking your doctor?
Some preparation makes the appointment more useful.
- Write down the date of your last seizure and any near misses, such as unexplained blank spells.
- Bring past EEG and MRI reports. They influence the relapse estimate.
- Ask what your own risk looks like, and what you'd do if a seizure returned, including driving and work.
- Keep sleep regular and take every dose until the plan changes. Sleep loss, stress and missed tablets are recognised seizure triggers.
If a seizure doesn't stop by itself, or one follows another without recovery, call 112.
What does this mean for children on levetiracetam or valproate?
Choosing between drugs is a separate question from stopping them. A 2026 meta-analysis of eight trials and 1,000 children found no clear difference in seizure freedom between levetiracetam and valproate, though the certainty was low. Levetiracetam was linked with fewer side effects overall and less weight gain, while behavioural and liver safety stayed uncertain [4]. In short, the choice has to fit the individual child.
| Group | Seizure-free period linked to higher relapse risk | Relative risk (95% CI) |
|---|---|---|
| All people with epilepsy | Under 5 years | 1.49 (1.17 to 1.88) |
| Children | Under 3 years | 1.70 (1.20 to 2.39) |
| Adults | Under 5 years | 1.52 (1.02 to 2.26) |
Seizure-free period and relapse risk after stopping anti-seizure medicine, from the 2026 meta-analysis [1].
Getting help in Delhi
At Brain Bridge Clinic in Safdarjung Enclave, New Delhi, we don't prescribe, change or stop anti-seizure medicine for epilepsy. That stays with your treating neurologist. What we offer is supportive care alongside it: a structured assessment of sleep, memory and attention, and how seizures affect daily life, with a sleep hygiene programme and cognitive rehabilitation where they're needed. At the first visit we go through your history and your neurologist's plan, then agree a few practical goals with you. If you'd like this kind of support, you can book an assessment. If you're feeling low or overwhelmed, Tele-MANAS (14416) is free, 24 hours a day.
Related pages: Epilepsy, Sleep Hygiene Program, Cognitive Rehabilitation Therapy.
Frequently asked questions
Can I stop epilepsy tablets by myself if I haven't had a seizure for years?
No. Stopping suddenly can bring seizures back, and the review found risk is higher after shorter seizure-free periods [1]. Decide with your neurologist, who can plan a gradual reduction if it's appropriate.
How many years seizure-free is enough?
It depends on your age and epilepsy. In this review, relapse risk was higher with under 3 years in children and under 5 years in adults [1]. Your neurologist looks at your whole picture, not one number.
Will my child need medicine for life?
Not necessarily. In one long follow-up, 77% of children treated for epilepsy of unknown cause were seizure-free without medicine decades later [2]. That doesn't apply to every type of epilepsy.
Is a slow taper safer than a fast one?
A small trial found similar relapse numbers with a 60-day and a 160-day taper, but it was too small to be conclusive [3]. Follow the schedule your doctor gives you.
Does Brain Bridge treat seizures?
No. We provide supportive care for sleep, thinking skills and quality of life alongside your neurologist's plan. We don't change seizure medicine.
When is a seizure an emergency?
Call 112 if a seizure doesn't stop on its own, if seizures repeat without recovery between them, or if the person is injured or has trouble breathing.
References
- Zhou H, Zhou W, Xu X. The effect of seizure-free period on seizure recurrence following antiseizure medication withdrawal: a systematic review and meta-analysis. Journal of neurology. 2026;273(11). doi:10.1007/s00415-026-14217-x. PMID 42900237.
- Braathen G, Zelano J. Long-term seizure and psychosocial outcome in childhood epilepsy of unknown cause. Acta neurologica Scandinavica. 2021;143(6):653-660. doi:10.1111/ane.13414. PMID 33772757.
- Ferlazzo E, Giussani G, Gasparini S, et al. Rapid versus slow withdrawal of antiepileptic monotherapy in two-year seizure-free adults patients with epilepsy (RASLOW) study: A pragmatic multicentre, prospective, randomized, controlled study. Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology. 2022;43(8):5133-5141. doi:10.1007/s10072-022-06121-9. PMID 35648267.
- Alharran AM, Alsaleeli KM, Alkhamis AS, et al. Levetiracetam versus valproate for new-onset or previously treated but inadequately controlled epilepsy in children: a systematic review and meta-analysis of randomized controlled trials. European journal of clinical pharmacology. 2026;82(11). doi:10.1007/s00228-026-04201-9. PMID 42900391.