PURPOSE: To investigate the predictive value of overgeneral memory (OGM) for outcome of electroconvulsive therapy (ECT) for depression. METHODS: The Autobiographical Memory Test was used to measure OGM in 25 patients with depression before ECT. The Hamilton Rating Scale for Depression (HRSD) was administered weekly to 1 week posttreatment. RESULTS: Overgeneral memory did not predict HRSD scores from the last ECT treatment, but did predict HRSD change scores from the last treatment to 1-week follow-up: patients high in OGM experienced a relatively greater increase in HRSD scores after the last treatment. CONCLUSIONS: Results further extend the status of OGM as a predictor of an unfavorable course of depression to a previously unstudied ECT population.