Background Superior cognitive functioning for electroconvulsive therapy (ECT) with right unilateral (RUL) ultrabrief pulse (UBP) stimulation compared to RUL brief pulse (BP) stimulation is not clearly established and long-term data is needed. Methods We conducted a prospective naturalistic follow-up of 87 inpatients from three tertiary psychiatric hospitals. Before these patients entered the follow up phase, they had participated in a RCT comparing twice weekly RUL BP (1.0 ms) with RUL UBP (0.3-0.4 ms) ECT eight times seizure threshold until remission (MADRS<10), for a maximum of six weeks. Three and six months after the index ECT patients were monitored for relapse and cognitive performance (retrograde amnesia, semantic memory and lexical memory). We compared relapse rate and cognitive performance between RUL BP and RUL UBP stimulation. Results Of the 50 patients who remitted after index ECT 44 (24 BP; 20 UBP) were monitored for follow up. Relapse occurred in 25% of the BP group and in 25% of the UBP group (χ2=0.00, p=1.0) at three-month follow-up; whereas 43.5% of the BP group and 35% of the UBP group relapsed (χ2=0.322, p=0.57) at six months follow-up. Cognitive assessments (17 BP; 16 UBP) showed no significant differences between BP and UBP groups, except for an advantage for the BP group in the autobiographical incident questions at three months follow-up only (p=0.04; d=0.77). Limitations This study may be limited since relapse in a naturalistic follow-up can be influenced by medication and other unknown factors, like social support, medical comorbidity, and psychotherapy. The small numbers of our subgroups hamper statistical significance. Conclusions Patients that achieved remission after RUL BP or RUL UBP ECT showed similar relapse rates after three and six months. There was no cognitive advantage of UBP over BP ECT in follow up. Clinical trials registration Netherlands trial register www.trialregister.nl registration number NTR1304.