The seizure threshold to electroconvulsive therapy (ECT) is defined entirely by the duration of the induced seizure, is multidetermined, and varies enormously with a wide variety of patient and treatment factors, including especially the parameters of the ECT stimulus. No consistent relationship has ever been detected between the clinical antidepressant response to ECT and either the threshold or the duration of the induced seizure.The stimulus titration method for determining the seizure threshold (titration–threshold dosing) was the central research tool used to reverse years of dogma by proving that the induced seizure of ECT is not alone sufficient to explain the therapeutic properties of ECT, and that the interaction between dosage and treatment electrode placement is critical in this regard. In clinical use, however, titration–threshold dosing has proven less than fully effective in optimizing the stimulus dose for ECT—better results are consistently obtained with age-based or fixed high-dose methods.The lack of a direct correlation between either seizure threshold or duration and clinical ECT response is an irremediable flaw of the titration–threshold method as clinically applied. New approaches are now called for in which ECT stimulus dosage is set and adjusted (“titrated”) according to the clinical antidepressant response of the patient or to measurable correlates thereof: maximum sustained electroencephalogram (EEG) ictal power, EEG postictal suppression, induced interictal EEG delta activity, peak heart rate, maximum sustained EEG coherence, and postictal EEG coherence reduction, all of which have been found by various investigators to be related to the clinical antidepressant response to ECT.