The question of whether transvenous or epicardial implantable cardioverter defibrillator (ICD) system is more beneficial in children and adolescents is controversially discussed. We sought to analyse the long-term outcome after implantation of ICDs using epicardial pacing/sensing and pleural shock leads. Retrospective analysis of 31 consecutive patients undergoing a total of 55 implantations of epicardial/pleural ICD systems below 20 years of age. Median age at implantation was 11.4 years (range 2.2-20) and median follow-up 57 months (range 0.4-127). The ICDs were implanted for primary (n = 17) and for secondary prevention (n = 14). The first defibrillation threshold at implantation was ≤25 J in 94% of the implant procedures. Appropriate shocks occurred in 6 of 31 patients. Inappropriate shocks occurred in 4 of 31 patients triggered by lead fracture, T-wave oversensing, sinus tachycardia, and atrial fibrillation. Freedom from first ICD discharge was 81, 71, and 71% at 3, 6, and 9 years, respectively. Reoperation was indicated in 16 of 31 patients for lead failure (n=11), end of battery life (n=10), generator migration (n=1) and recall (n=1); freedom from reoperation was 74 and 55% at 3 and 6 years. Paediatric epicardial/pleural ICD therapy is feasible, effective, and safe both in the short-term as well as in the long-term perspective at the price of relatively frequent surgical revisions. They do not generally result in an increased burden of inappropriate shocks. This ICD system meets the needs of the paediatric population and can be recommended as a first choice in this age group.