The upper limit of vulnerability is the strength above which ventricular fibrillation is no longer inducible with a shock delivered during the vulnerable phase of the cardiac cycle. It has been demonstrated that the upper limit of vulnerability correlates with the defibrillation threshold in a paced rhythm. The purpose of this study is to evaluate the correlation of the upper limit of vulnerability determined in normal sinus rhythm with the defibrillation threshold using a simplified protocol in patients undergoing placement of an ICD. We studied 28 patients who underwent ICD implantation. CPI generators and Endotak leads were used in all patients. Device-based testing was used to determined the defibrillation threshold and the upper limit of vulnerability. The upper limit of vulnerability was tested with three shocks delivered at 0, 20, and 40 ms before the peak of the T wave during normal sinus rhythm. The defibrillation threshold was determined by a simple step up-down protocol. The upper limit of vulnerability (9.0 +/- 4.5 J) did not significantly differ from the defibrillation threshold (9.9 +/- 4.0 J), P = NS. A close correlation was present, correlation coefficient = 0.75, P < 0.0001. The upper limit of vulnerability was within 5 J of the defibrillation threshold in 27 (96%) of the 28 patients. The upper limit of vulnerability underestimated the defibrillation threshold by 10 J in one patient who had a defibrillation threshold of 15 J. The upper limit of vulnerability determined in normal sinus rhythm correlates significantly with the defibrillation threshold in patients undergoing ICD implantation. The protocol is simple and easily implemented clinically.