Objectives. Electroconvulsive therapy (ECT) is an effective treatment for severe depression. Electroencephalogram (EEG) measures between ECT sessions seem to be related to the antidepressant efficacy of ECT. In this naturalistic cohort study, we examine longitudinal effects of ECT on interhemispheric EEG coherence measures during seizure activity and its relation to the antidepressant efficacy. Methods. This study included 65 patients diagnosed with severe depressive disorder. Depressive symptoms were rated according to the Montgomery-Åsberg Depression Rating Scale before and after the course of ECT. Frequency-specific ictal interhemispheric (fp1-fp2) EEG coherence measures were established during the first and each consecutive sixth treatment session. Linear mixed-effect models were used to determine longitudinal changes in ictal coherence measures during the course of ECT and its relation to treatment efficacy. Results. Ictal interhemispheric coherence in the theta and alpha frequency bands increased over the course of treatment, whereas no significant change was found for the delta and beta frequency bands. A main effect of treatment efficacy on the interhemispheric coherence in the delta and theta band was revealed. However, the longitudinal effects of ECT were not associated with treatment efficacy. Conclusion. The current study suggests that interhemispheric coherence during ECT-induced seizures increases over the course of treatment. Furthermore, these longitudinal effects seem to be unrelated to the antidepressant efficacy of ECT. These findings contribute to the understanding of the mechanism of action of ECT.