PurposeThere is limited research on the proportion of individuals with epilepsy who maintain response to ketogenic diet therapy (KDT) after discontinuing treatment. We aimed to determine the proportion of individuals who did / did not maintain response post KDT and explore factors that may influence the likelihood of maintaining response. MethodsRetrospective data were collected from 97 individuals from 9 KDT centres. Individuals had achieved ≥50% seizure reduction on KDT for at least 12 months, with seizure frequency data available at 3 months+ post diet. Outcome 1 was: recurrence of seizures or increase in seizure frequency post diet; outcome 2: recurrence of seizures, increase in seizure frequency or an additional anti-seizure treatment started post diet. Results61/97 (62.9%) individuals maintained response at latest follow-up (mean 2.5[2.0] years since stopping KDT). Approximately one third maintained response without further anti-seizure treatments. One quarter of individuals had an increase in frequency or recurrence of seizures within 6 months (95%CI 4, 12) for outcome 1 and within 3 months (3, 6) for outcome 2.Individuals who did not achieve seizure freedom on diet were significantly more likely to have an increase in seizures or to require additional anti-seizure treatments post diet compared to those who were seizure-free on diet (hazard ratio 4.02, 95%CI (1.46, 11.16) p < 0.01). ConclusionOur findings should help guide clinical teams with the information they provide patients and their families regarding likelihood of long-term seizure response to KDT. Realistic costings for KDT services may need to be considered.