The retrograde amnesia following electro-convulsive treatment was studied by means of a memory test battery in 3 equivalent groups of patients, each comprising 24 patients. In one group, treatments were given with ultrabrief stimulus technique (Either electroshock apparatus type ES), in the second (the control group with standard technique) with pulses of longer duration (Siemens Konvulsator III) and in the third with the latter apparatus and, in addition, with reduction of the seizure discharge by lidocaine. 1) With ultrabrief stimulus technique, seizures could be induced with a reduced amount of electric energy. The seizures were of the same duration as in standard treatment but 4 patients presented deviating features comprising breathing before the end of the seizure and rapid post-shock recovery. Retrograde amnesia was lower than in the control group. 2) In the group where lidocaine had been given, the electric-energy applied was somewhat higher than in the control group. Seizure duration was considerably reduced. Retrograde amnesia was of the same level as that in the control group. The observations may be in line with the view that the memory disturbance after ECT is partly a direct effect of the current and is related to the amount of energy applied. Due to the deviating features of some seizures induced with ultrabrief stimuli, however, the possibility that the reduced memory disturbance has arisen via a change of the seizure cannot be excluded.