There is an ongoing debate as to whether propofol exhibits pro- or anticonvulsant effects, and whether it should be used in patients with epilepsy. We prospectively assessed the occurrence of seizure-like phenomena and the effects of intravenous propofol on the electroencephalogram (EEG) in 25 children with epilepsy (mean (SD) age: 101 (49) months) and 25 children with learning difficulties (mean (SD) age: 52 (40) months) undergoing elective sedation for MRI studies of the brain. No child demonstrated seizure-like phenomena of epileptic origin during and after propofol sedation. Immediately after stopping propofol, characteristic EEG changes in the epilepsy group consisted of increased beta wave activity (23/25 children), and suppression of pre-existing theta rhythms (11/16 children). In addition, 16 of 18 children with epilepsy and documented EEG seizure activity demonstrated suppression of spike-wave patterns after propofol sedation. In all 25 children with learning difficulties an increase in beta wave activity was seen. Suppression of theta rhythms occurred in 11 of 12 children at the end of the MRI study. In no child of either group was a primary occurrence or an increase in spike-wave patterns seen following propofol administration. The occurrence of beta wave activity (children with learning difficulties and epilepsy group) and suppression of spike-wave patterns (epilepsy group) were transient, and disappeared after 4 h. This study demonstrates characteristic, time-dependent EEG patterns induced by propofol in children with epilepsy and learning difficulties. Our data support the concept of propofol being a sedative-hypnotic agent with anticonvulsant properties as shown by depression of spike-wave patterns in children with epilepsy and by the absence of seizure-like phenomena of epileptic origin.