Defibrillation threshold (DFT) testing is performed routinely in patients undergoing implantable cardioverter-defibrillator (ICD) implantation to verify the ability of the ICD to terminate ventricular fibrillation (VF). However, neither the efficacy nor the safety of DFT testing has been proven; thus, the necessity of such testing is controversial. We conducted a retrospective study of the efficacy of DFT testing, particularly with respect to long-term outcomes of ICD implantation.The study included 150 patients (125 men, 25 women, aged 59.0 ± 17.6 years) who underwent ICD or cardiac resynchronization therapy defibrillator implantation, with (n = 73) or without (n = 77) intraoperative DFT testing, between June 1996 and September 2007. VF was induced by delivery of a T-wave shock, and a 20-25-J shock was then delivered. If the 20-25-J shock failed to terminate VF, 30 J was delivered. We assessed whether undersensed VF events occurred during DFT testing and/or during patient follow-up and checked for any association between undersensing and delayed shock delivery. During DFT testing, fine VF was sensed, and shocks were delivered in a timely manner. Nevertheless, 2 patients in the DFT testing group died from VF within 3 years after device implantation.DFT testing, in comparison to non-DFT testing, appeared to have no influence on the long-term outcomes of our patients, suggesting that DFT testing at the time of ICD implantation is limited.