OBJECTIVE Currently, there continues to be a lack of evidence regarding outcomes associated with device-based therapy for ventricular arrhythmias in elderly patients, even more in primary-prevention indications. We aimed to describe the follow-up in terms of efficacy and safety of implantable cardioverter-defibrillator (ICD) therapy in a large cohort of elderly patients. RESULTS Retrospective multicentre study performed in 15 Spanish hospitals. Consecutive patients referred for ICD implantation before 2011 were included. One hundred and sixty-two of 1174 patients (13.8%) ≥75 years were considered as 'elderly'. When compared with those patients <75, this subgroup presented more co-morbid conditions, including hypertension, chronic obstructive pulmonary disease , and renal failure, and more previous hospitalizations due to heart failure (HF). During a mean follow-up of 104.4 ± 3.3 months, 162 patients (14%) died, 120 in the younger age (12.4%), and 42 (24.4%) in the elderly. Kaplan-Meier analysis showed an increased probability of death with increasing age (17, 24, 28, and 69% at 12, 24, 48, and 60 months of follow-up in the elderly group). There was neither difference regarding the rate of appropriate nor inappropriate ICD intervention. CONCLUSIONS In a real-world scenario, elderly patients comprise ∼15% of ICD implantations for primary prevention of sudden cardiac death (SCD). Although the rate of appropriate therapy is similar between groups, the benefit of ICD is attenuated for a major increase in mortality risk among those patients ≥75 years at the moment of device implantation.