OBJECTIVE: To assess the safety and efficacy of an early home-based walking program for first-time implantable cardioverter-defibrillator (ICD) recipients. DESIGN: Pre-post intervention trial. SETTING: Institutional and private practice. PARTICIPANTS: Cardiac patients (N=301) with an initial ICD implantation for primary or secondary prevention; able to read, speak, and write English; and having access to telephone. INTERVENTIONS: Early home-based walking protocol implemented 1-month post-ICD implant. Exercise tolerance monitored by study nurses via telephone. MAIN OUTCOME MEASURES: Safety assessment was based on the frequency of ICD therapies and hospitalizations, and efficacy assessment was based on pedometer measures and self-report of ICD self-efficacy and physical activity. RESULTS: ICD recipients were on average 64.1±11.9 years old, predominantly men, and white, with an ejection fraction of <35% and a mean Charlson comorbidity score of 2.3±1.5. Nineteen individuals (6.3%) received 28 ICD shocks; 15 (53.6%) were appropriate and 13 (46.4%) inappropriate. Antitachycardia pacing therapies were delivered 72 times in 18 individuals (6%), with 61 (84.7%) being appropriate and 11 (15.3%) inappropriate. Five ICD shocks (2 appropriate and 3 inappropriate) and 2 antitachycardia pacing therapies occurred during walking. Five participants (2%) were hospitalized for an ICD shock, none of which was associated with walking. Average steps per day increased by 806 over 3 months. Perceived exercise self-efficacy improved significantly as did weekly exercise. Predictors of receiving any ICD shock were younger age (P<.0001), moderate to severe renal disease (P=.001), and lymphoma (P=.024). CONCLUSIONS: Early ambulation after an initial ICD was safe and effective, with few ICD shocks and improved efficacy.