OBJECTIVES: This study sought to identify the incidence and outcome related to inappropriate implantable cardioverter-defibrillator (ICD) shocks, that is, those for nonventricular arrhythmias. BACKGROUND: The MADIT (Multicenter Automatic Defibrillator Implantation Trial) II showed that prophylactic ICD implantation improves survival in post-myocardial infarction patients with reduced ejection fraction. Inappropriate ICD shocks are common adverse consequences that may impair quality of life. METHODS: Stored ICD electrograms from all shock episodes were adjudicated centrally. An inappropriate shock episode was defined as an episode during which 1 or more inappropriate shocks occurred; another inappropriate ICD episode occurring within 5 min was not counted. Programmed parameters for patients with and without inappropriate shocks were compared. RESULTS: One or more inappropriate shocks occurred in 83 (11.5%) of the 719 MADIT II ICD patients. Inappropriate shock episodes constituted 184 of the 590 total shock episodes (31.2%). Smoking, prior atrial fibrillation, diastolic hypertension, and antecedent appropriate shock predicted inappropriate shock occurrence. Atrial fibrillation was the most common trigger for inappropriate shock (44%), followed by supraventricular tachycardia (36%), and then abnormal sensing (20%). The stability detection algorithm was programmed less frequently in patients receiving inappropriate shocks (17% vs. 36%, p = 0.030), whereas other programming parameters did not differ significantly from those without inappropriate shocks. Importantly, patients with inappropriate shocks had a greater likelihood of all-cause mortality in follow-up (hazard ratio 2.29, p = 0.025). CONCLUSIONS: Inappropriate ICD shocks occurred commonly in the MADIT II study, and were associated with increased risk of all-cause mortality.