Defibrillation is thought to be mediated by a depolarizing current; however, the present method of defibrillation is based on delivering an empiric dose of energy to all patients. The hypothesis of this study was that for equivalent efficacy rates, a current-based defibrillation method would result in delivering less energy and peak current than would the standard energy-based method. In a group of 86 consecutive patients with ventricular fibrillation, every other patient was prospectively assigned to receive shocks according to method 1 or method 2. Method 1 was current based and delivered successive shocks of 25, 25 and a maximum of 40 A; method 2 was energy based and delivered shocks of 200, 200 and 360 joules. Patients in both groups were similar with respect to age, gender, weight, cardiac diagnosis, ejection fraction, antiarrhythmic therapy, chest circumference, chest depth and transthoracic impedance. Each method had statistically equivalent first shock (79% current-based versus 81% energy-based) and cumulative shock success rates. The mean first shock energy was 120 +/- 30 joules for patients receiving the current-based method and 200 joules for patients receiving energy-based shocks (p = 0.0001). The mean peak current was 24 +/- 2.3 and 33 +/- 5.0 A, respectively (p = 0.0001). Therefore, for equivalent first shock success rates, the energy-based method delivered 67% more energy and 38% more current than the current-based method. High transthoracic impedance (greater than or equal to 90 omega) predicted first shock failure only in patients undergoing defibrillation by the energy-based method (p = 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)