The theoretical tissue model-based estimates of phase 1 and phase 2 duration of biphasic waveforms are considerably shorter than the pulse widths currently used in ICDs with standard tilt. This study used a tissue resistance/capacitance (RC) model to identify optimal biphasic pulse widths. By paired step-down defibrillation threshold (DFT) testing, the efficacy of standard versus "tuned" biphasic waveforms was evaluated in 91 patients. Standard waveforms consisted of a phase 1 set to 65% tilt and phase 2 = phase 1. The tuned waveform was based on an RC model of membrane characteristics with a time constant of 3.5 ms. The optimal phase 1 truncation point is at the peak of membrane response. The optimal phase 2 duration ends with a membrane response near or just below 0. In paired analysis, no significant differences were found in DFT or impedance between standard and tuned waveforms. In patients with DFTs > 400 V, the tuned waveform lowered the DFT by an average of 38 V (P < 0.05). Multivariate analyses showed a significant inverse relationship between DFT and impedance (P < 0.001). As impedance increased, the tuned waveform was associated with DFTs comparable to the standard waveform with shorter pulse duration and lower delivered energy. No single tilt value allowing an easy calculation of delivered energy was related to ICD waveform efficacy. The use of ICDs with tuned optimal pulse durations offer a greater flexibility of choice for patients with high DFTs.