Psychomotor symptoms are core features of melancholic depression. This study investigates whether psychomotor disturbance predicts the outcome of electroconvulsive therapy (ECT) and how the treatment modulates psychomotor disturbance. In 73 adults suffering from major depressive disorder psychomotor functioning was evaluated before, during and after ECT using the observer-rated CORE measure and objective measures including accelerometry and a drawing task. Regression models were fitted to assess the predictive value of melancholic depression (CORE ≥ 8) and the psychomotor variables on ECT outcome, while effects on psychomotor functioning were evaluated through linear mixed models. Patients with CORE-defined melancholic depression (n = 41) had a 4.9 times greater chance of reaching response than those (n = 24) with non-melancholic depression (Chi-Square = 7.5, P = 0.006). At baseline, both higher total CORE scores (AUC = 0.76; P = 0.001) and needing more cognitive (AUC = 0.78; P = 0.001) and motor time (AUC = 0.76; P = 0.003) on the drawing task corresponded to superior ECT outcomes, as did lower daytime activity levels (AUC = 0.76) although not significantly so after Bonferroni correction for multiple testing. A greater CORE-score reduction in the first week of ECT was associated with higher ECT effectiveness. ECT reduced CORE-assessed psychomotor symptoms and improved activity levels only in those patients showing the severer baseline retardation. Although the sample was relatively small, psychomotor symptoms were clearly associated with beneficial outcome of ECT in patients with major depression, indicating that monitoring psychomotor deficits can help personalise treatment.