Objectives. This study was performed to test the hypothesis that sodium channel activity is important in determining ventricular vulnerability to electric shocks. Background. It is unknown whether sodium channel activity determines the upper limit of vulnerability. Methods. The ventricles of 10 open chest dogs were paced at 300-, 500-, or 1,000-ms cycle lengths. The shock strength associated with a 50% probability of reaching the upper limit of vulnerability (ULV(50)) and the shock strength associated with a 50% probability of defibrillation (DFT50) were then determined by means of an up-down algorithm. Lidocaine (9.2 mg/kg body weight loading dose and 285-mu g/kg per min maintenance dose) was then given, and the ULV(50) and the DFT50 were redetermined after 1 h of stable infusion. Results. The mean (+/- SD) lidocaine concentration was 11.9 +/- 2.4 mu g/ml. At baseline, the ULV(50) tested with each S-1 cycle length was not significantly different from the DFT50. During lidocaine infusion, the ULV(50) determined with cycle lengths of 300 and 500 ms (18.9 +/- 11.3 and 16.1 +/- 8.9 J, respectively) were significantly (p < 0.05) higher than those simultaneously determined for the DFT50 (11.2 +/- 4.1 and 10.9 +/- 5.6 J, respectively). However, when determined with an S-1 cycle length of 1,000 ms, the ULV(50) (10.4 +/- 4.1 J) was not found to be significantly different from the DFT50 (10.3 +/- 5.3 J). Lidocaine infusion increased (p < 0.05) QRS duration and the effective refractory periods for cycle lengths of 300 and 500 ms but not 1,000 ms. Conclusions. The effect of lidocaine on the upper limit of vulnerability is use dependent. These results are compatible with the hypothesis that sodium channel activity is important in determining ventricular vulnerability to electrical shocks.