Determination of an adequate defibrillation safety margin (DSM) or defibrillation threshold (DFT) can sometimes be a vexing problem in patients with implantable defibrillators. The upper limit of vulnerability has been shown to be highly correlative with measured defibrillation thresholds. This technique has been met with resistance by many implanting electrophysiologists as determination of the vulnerable period of the cardiac cycle may require numerous high energy shocks and can be very time consuming. The peak of the latest peaking monophasic T wave measured from the pacing spike of the induction train of S1 via the surface ECG is generally accepted as the vulnerable period of the cardiac cycle. This peak, however, serves as the starting point for delivery of shocks to determine the upper limit of vulnerability and requires three to four shocks per energy level. The coupling interval measured from the pacing spike to the T wave peak is generally the interval for the first test shock. Subsequent shocks are given at -20 ms, +20 ms, and +40 ms relative to the peak of the T wave. The purpose of this study was to determine if a single IEGM derived coupling interval could provide an accurate determination of the vulnerable period to utilize a modified upper limit of vulnerability test method for delineating a defibrillation safety margin. A total of 72 patients that indicated for single, dual chamber or CRT-D implantable defibrillators were included in the study. Wide fluctuations in the IEGM and surface ECG derived coupling intervals were present and this difference met statistical significance, t(71) = 4.87, p <.001. A strong negative correlation was achieved, r (72) = -.909, p < .001, indicating that as the modified upper limit of vulnerability went up, the defibrillation safety margin went down. Partial correlation failed to reveal any association between the MULV and the IEGM coupling interval indicating that this is a patient specific physiologic measurement that varies from patient to patient. Results of the regression analysis predicting IEGM coupling intervals from the various predictors were also found to be significant at .01. F (14, 57) = 2.53, p = .007. Finally, the model that a T shock at or above the ULV that does not induce VF would defibrillate VF 90% of the time was satisfied in 69 of 72 (96%) patients. Because of the high correlation between ULV and DFT, a modified ULV may be utilized as a surrogate for an estimation of the DFT.