The aim of this study was to evaluate ventricular arrhythmias occurring in recipients of the InSync ICD for the primary and secondary prevention of sudden death. The InSync ICD was implanted in 142 patients (128 men; mean age 65 +/- 10 years) with heart failure (mean NYHA functional Class 3.0 +/- 0.7) and wide QRS (mean 159 +/- 33 ms). The underlying etiology was ischemic in 55%, idiopathic in 33%, and valvular or hypertensive cardiomyopathy in 12% of patients. The numbers of arrhythmic episodes/100 patient-months was computed with their 95% CI, assuming a Poisson distribution. Implants were performed in 48 (34%) patients who did not have an ACC/AHA guidelines Class I indication for ICD therapy. A total of 104 patients were compliant for follow-up visits. During a 9-month median (range 0.1-24) follow-up of 104 compliant patients, 19 experienced a total of 94 ventricular arrhythmias, all successfully interrupted or self-terminated, with a median number of two separate episodes, corresponding to a rate of 10 episodes/100 person-month (95% CI 8-12). A rate of 12 episodes/100 person-months (95% CI 10-15) was measured in the subgroup of patients with ACC/AHA class I indications, versus two episodes/100 person-months (95% CI 1-5) in the remainder of the population. Among 12 deaths, 9 were due to heart failure, 1 to a non-cardiovascular cause, and 2 to unknown causes. The implantation of ICD in heart failure patients has been prominently extended to primary prevention. Patients without standard ICD indications experienced life-threatening arrhythmic events. The impact of ICD combined with cardiac resynchronization therapy on arrhythmic profile, mortality, and costs in this subgroup of patients need to be more precisely studied, with a particular focus on the various types of underlying heart disease.