Objectives Electroconvulsive therapy (ECT) is used in some cases of postpartum depression (PPD) and postpartum psychosis (PPP). The risk of relapse for PPD and PPP after ECT is unknown. This study compared the relapse rate after ECT between women who had been treated for PPD and/or PPP and women who had been treated for depression and/or psychosis outside the postpartum period. Methods The Swedish National Quality Register for ECT and the Swedish National Patient Register were used to identify women with PPD and/or PPP who had been treated with ECT within 6 months after delivery. For each case, a control (treated with ECT but not postpartum) patient was also selected. A Kaplan-Meier estimator was used to calculate the relapse rate (defined as rehospitalization or suicide) after ECT. Cox regression was used to identify variables associated with relapse. Results A total of 180 patients were included in each group. The proportions of patients who suffered relapse after 6 months, 1 year, and 2 years were 28%, 31%, and 40% for the postpartum group and 39%, 50%, and 55% for the nonpostpartum group. Treatment with benzodiazepines, several previous psychiatric admissions, and the absence of improvement after ECT were associated with relapse. Conclusions The risk of relapse after ECT is lower for patients with PPD and/or PPP than for patients outside the postpartum period, but the risk is nonetheless substantial in both groups.