In a hospital-based study, 494 children with epilepsy were prospectively followed up from the time of diagnosis. The main objective of this study was to investigate the course of childhood-onset epilepsy during a period of 15 years. Generally, medication is withdrawn after a 2-year remission. Here, subjects with a fast response to medication were randomized for 6- or 12 month treatment. Four years later, we found no difference in recurrence rate, terminal remission, or adverse events between both groups. Unnecessary treatment was prevented for >1 year, but after recurrences renewed medication sometimes failed. We cannot recommend early withdrawal in all fast responders. At last contact, a 5-year terminal remission was reached by 71% of the cohort and 9% was intractable. Course during follow-up was favorable in 50%, improving in 29%, and poor or deteriorating in 16%. Duration of seizure activity was <1 year in 25% of the cohort and >12 years in another 25%. Antiepileptic drugs were used by 86%. Mortality was significantly higher only in the remote symptomatic group. The idiopathic and cryptogenic etiology groups resembled their Dutch age-peers in having a partner/ offspring and in employment status, but the educational and occupational achievements of the idiopathic group were lower than expected. This might be due to their epilepsy, or to an assumed common cause for the epilepsy and other brain dysfunctions. Time of onset, course, and duration of intractability can differ between subjects with epilepsy and are hardly predictable, even though intractability itself can be predicted to a certain extent. The natural course of epilepsy probably best explains the variability of intractability and the effect of medication seems to be minor.