BACKGROUND: Electroconvulsive therapy (ECT) is an effective treatment for patients with severe depression. Knowledge on factors predicting therapeutic response may help to identify patients who will benefit most from the intervention. Based on the neuroplasticity hypothesis, volumes of the amygdala and hippocampus are possible candidates for predicting treatment outcome. Therefore, this prospective cohort study examines the predictive value of amygdala and hippocampal volumes for the effectiveness of ECT. METHODS: Prior to ECT, 53 severely unipolar depressed patients [mean age 57 +/- 14 years; 40% (n = 21) male] received structural magnetic resonance imaging (MRI) at 1.5 T. Normalized amygdala and hippocampal volumes were calculated based on automatic segmentation by FreeSurfer (FS). Regression analyses were used to test if the normalized volumes could predict the response to a course of ECT based on the Montgomery-Asberg Depression Rating Scale (MADRS) scores. RESULTS: A larger amygdala volume independently and significantly predicted a lower post-ECT MADRS score (beta = -0.347, P = 0.013). The left amygdala volume had greater predictive value for treatment outcome relative to the right amygdala volume. Hippocampal volume had no independent predictive value. CONCLUSION: A larger pre-treatment amygdala volume predicted more effective ECT, independent of other known predictors. Almost all patients continued their medication during the study, which might have influenced the course of treatment in ways that were not taken into account.