U ntil 2015, consensus statements 1,2 advised against sports participation more vigorous than golf for patients with implantable cardioverter-defibrillators (ICDs) because of the postulated risks of death caused by failure to defibrillate, injury resulting from arrhythmia-related syncope or shock, or device damage. The multinational, prospective, observational ICD Sports Safety Registry quantified risks associated with sports participation for athletes receiving ICDs on the basis of standard criteria. Initial results (2013 3 ) demonstrated no death, failure to defibrillate, or injury resulting from arrhythmia or shock during sports. On the basis of these data, the 2015 eligibility and disqualification recommendations for competitive athletes with cardiovascular disease 4 now state that competitive sports may be considered for athletes with ICDs. This report describes 4-year follow-up of the completed registry. Methods are as reported previously.3 The Yale University Human Investigation Committee approved the study. All participants gave written informed consent.Among 440 participants, 393 in organized sports and 47 in high-risk sports, the most common diagnoses were long-QT syndrome (n=87, 20%), hypertrophic cardiomyopathy (n=75, 17%), and arrhythmogenic right ventricular cardiomyopathy (n=55, 13%). Of 201 subjects with a preimplantation history of ventricular fibrillation (VF) or tachycardia (VT), 61 (30%) had VT/VF during sports. At enrollment, median time since implantation was 26 months (interquartile range, 11-59 months), with 126 subjects (29%) enrolled within 1 year of implantation. The most common organized sports were running, basketball, and soccer; the most common dangerous sport was skiing. Seventy-seven subjects (18%) engaged in varsity/junior varsity/ traveling team competition, (highly competitive subgroup). Seventy-two postcollege athletes (16%) participated at a national/international level.Median follow-up was 44 months (interquartile range, 30-48 months), totaling 1446 person-years. Thirty-seven participants did not complete the study: 20 were lost to follow-up (all confirmed alive), 5 withdrew, 6 developed worsening cardiac/ medical conditions, 4 had the ICD removed, and 2 died (neither death was sports related, as reported previously 3 ). There were no tachyarrhythmic deaths or externally resuscitated tachyarrhythmias during or after sports participation or injury resulting from arrhythmia-related syncope or shock during sports. The 95% confidence interval for the occurrence of adverse event based on 376 participants followed up at least 2 years was 0% to 0.9% and based on 167 participants followed up at least 4 years was 0% to 2.2%.The numbers and rhythms of shocks received for the overall group and the highly competitive subgroup are shown in the Table. Forty-six (10%) received appropriate shocks (for VT/VF) during competition or practice, a rate of 3 per 100 person-years (identical to the initial report 3 ). More participants received shocks during competition/ practice or physical a...