EEG and EMG were monitored simultaneously at 72 ECT sessions in 18 patients. When EMG was recorded from the leg the duration of EMG curve convulsions varied between 43 and 89% of the EEG recorded seizure duration of the same fit. The EMG/EEG seizure ratio varied widely among different patients as well as within the single patient during a series of ECT sessions. EMG recording from the masseter muscles showed EMG seizure duration to be more in accordance with the EEG seizure duration than EMG recordings from the leg. The EMG/EEG seizure ratio (masseter) varied from 65 to 100%, which was significantly higher than the EMG/EEG seizure ratio obtained with EMG recorded from the calf muscles. The use of EEG monitoring of ECT is advocated in order to obtain the most precise evaluation of the cerebral seizure activity. If EEG monitoring is not available EMG recorded from the masseter muscles provides a reasonably accurate and easily applicable method of monitoring the presence and duration of the ECT-induced seizure.