OBJECTIVE: To describe the natural history of electroencephalography (EEG) changes in patients with benign epilepsy with centrotemporal spikes (BECTS) over 1 year. METHODS: Centrotemporal spikes were visually evaluated based on 24-h ambulatory EEG studies to determine the total, left, right, and bilateral centrotemporal spikes patients were awake and asleep. These spike rates were then used to compare the entire night of sleep to the first 2 h of sleep, the repeatability of spike frequency over two recordings (done within days to weeks), and longitudinal changes in spike rate over 6 and 12 months. RESULTS: Nineteen children with newly diagnosed and untreated BECTS were included in this analysis. An excellent correlation was found between the centrotemporal spike rate during the entire duration of sleep and the first 2 h of sleep (intraclass correlation [ICC] 0.87, 95% confidence interval [CI] 0.67-0.95). In addition, an excellent correlation was found between two recordings completed an average of 23 days apart while patients were asleep (ICC 0.92, 95% CI 0.80-0.97) and good correlation while patients were awake (ICC 0.70, 95% CI 0.39-0.87). The average change in spike rate between recordings at baseline and at 6 months was a decrease of 64.7% (range -100% to +51.5%, p = 0.01) and the average change in rate between recordings at baseline and at 12 months was a decrease of 57.7% (range -100% to +29.1%, p = 0.01). In addition, within 6 months, most children had decreased centrotemporal rates, with 30% of children being spike-free. This absence of spikes did not continue in all children, since the majority (60%) had some spikes at 1 year following diagnosis. SIGNIFICANCE: Centrotemporal spike rates during sleep are stable when compared over days to weeks; however, when comparing spike rates over months there is a larger degree of variability.