Background: Electroconvulsive therapy is an important somatic treatment for severe mental disorders with established efficacy and safety. However, data on the relationship between ECT and the readmission rate of patients with schizophrenia are scarce. This study will explore the association between the administration of ECT and readmission rates using a machine learning method. Methods: Inpatient medical records from the year of 2016 in one large psychiatric hospital in Beijing, China, were analyzed using a machine learning algorithm to determine the most important variables affecting readmission of patients with schizophrenia. Results: The medical records of 2131 inpatients with schizophrenia were reviewed. 1099 patients were followed up within 3 months of their index admission (642 ECT cases and 457 non-ECT cases) and 1032 patients were followed up within 6 months (596 ECT cases and 436 non-ECT cases) after discharge. The 3- and 6-month readmission rates in the ECT group (11.37% and 17.94%, respectively) were significantly lower than that of the patients who did not receive ECT (18.79% and 29.36%, respectively, both p