The objective of this study was to identify the factors, at the time of diagnosis, that determine the prognosis for remission of epilepsy. A prospective community‐based cohort study of 792 patients recruited at the time of their first diagnosis of epileptic seizures was undertaken; in those classified 6 months after presentation, the median follow‐up period was 7.2 years (quartiles at 6.2 and 8.2 years) after presentation. We analyzed data from 6 months after the first identified seizure, which prompted the dianosis of epilepsy, to allow us to factor in those aspects contingent on a diagnostic assessment. Baseline clinical and demographic data were analyzed using the Cox proportional hazards regression model with remission of epilepsy for 1, 2, 3, and 5 years as outcome measures. The dominant clinical feature predicting remission was the number of seizures in the 6‐month diagnostic assessment period. Thus, the chance of entering 1 year of remission by 6 years for a patient who had 2 seizures during this initial 6 months was 95%; for 5 years of remission, it was 47% as opposed to 75% for 1 year of remission and 24% for 5 years of remission if there had been 10 or more seizures during this period. The number of seizures in the early phase of epilepsy (here, taken as the first 6 months after presentation) is the single most important predictive factor for both early and long‐term remission of seizures. Ann Neurol 2000;48:833–841