BACKGROUND: Advances in implantable cardioverter defibrillator (ICD) technology have resulted in more effective defibrillation, and the need for defibrillation threshold (DFT) testing has been questioned. METHODS: A total of 64,227 initial ICD implant procedures performed at 1,261 facilities (April to December 2010) in the NCDR Registry™ were selected to identify patterns of practice related to DFT testing. Patient, physician, and procedural characteristics were compared. Independent association of DFT testing with in-hospital adverse events or mortality was also examined. RESULTS: DFT testing was performed in 71% of patients. Patients who did not undergo testing were older; more often had heart failure, lower left ventricular ejection fraction, atrial arrhythmias, and a primary prevention indication; and were more likely to receive a cardiac resynchronization therapy device. In addition, patients who did not undergo testing were less likely to be implanted in a metropolitan area and more often implanted at a teaching hospital, in New England, in the Mid-Atlantic, or in the Pacific U.S . In-hospital adverse events occurred in 2.56% of patients who underwent DFT testing compared to 3.58% who did not (P < 0.001). Death or any complication remained more likely to occur in patients who did not undergo testing (odds ratio and 95% confidence interval: 1.46 [1.33, 1.61], P < 0.001), after adjusting for baseline and procedural differences. CONCLUSIONS: DFT testing is not performed in many (29%) patients in clinical practice. Patients who did not undergo testing were more likely to have adverse events and to be older with more comorbidities, which could explain why DFT testing was avoided. Prospective randomized data are needed to determine the impact of DFT testing on outcome.