Introduction: Prompt defibrillation is critical during paediatric cardiac arrest. The main objective of this systematic review was to determine the initial defibrillation energy dose for ventricular fibrillation (VF) or pulseless ventricular tachycardia (pVT) that is associated with sustained return of spontaneous circulation (ROSC) during paediatric cardiac arrest. Associations between initial defibrillation energy dose with any ROSC, survival and defibrillation-induced complications were also assessed. Methods: A systematic review was performed using four databases (Medline, Embase, Web of Science, Cochrane Library) (PROSPERO: CRD42016036734). Human studies (cohort studies or controlled trials) and animal model studies (controlled trials) of pediatric cardiac arrest involving assessment of external defibrillation energy dosing were considered. The primary outcome was sustained ROSC. Two researchers independently reviewed all the titles and abstracts of the retrieved citations, selected the studies and extracted the data using a standardized template. Risk of bias of human non-randomised studies were assessed using the ROBIN-I tool (formerly ACROBAT-NRSI) tool proposed by the Cochrane Collaboration group. Results: The search strategy identified 14,471 citations of which 232 manuscripts were reviewed. Ten human and 10 animal model studies met the inclusion criteria. Human studies were prospective (n = 6) or retrospective (n = 4) cohort studies and included between 11 and 266 patients (median = 46 patients). Sustained ROSC rates ranged from 0 to 61% (n = 7). No studies reported a statistically significant association between the initial defibrillation energy dose and the rate of sustained ROSC (n = 7) or survival (n = 6). No human studies reported defibrillation-induced complications. Meta-analysis was not considered appropriate due to clinical heterogeneity. The overall risk of bias was moderate. All animal studies were randomized controlled trials with 8 and 52 (median = 27) piglets. ROSC was frequently achieved (more than 85%) with energy dose ranging from 2 to 7 joules/kg (n = 7). The defibrillation threshold varied according to the body weight and appears to be higher in infant models. Conclusion: Defibrillation energy doses and thresholds varied according to the body weight and trended higher for infants. No definitive association between initial defibrillation doses and the outcomes of sustained ROSC or survival could be demonstrated.