Projected pacemaker longevity is calculated according to the nominal setting, which is 3.5 V for pulse amplitude in most present day pacemakers. The aim of this study was to test whether the nominal ventricular output setting of 3.5 V and 0.4 ms provides a 100% safety margin if these pacemakers are connected to implanted ventricular pacing leads. The study included 24 patients undergoing pulse generator exchange. The new pacemaker was either a Thera DR (n = 21) or Thera SR (n = 3) device. Ventricular pacing leads were implanted 70 +/- 38 months previously. Intraoperative measurements included pacing threshold at 0.5-ms pulse duration, impedance, and R wave amplitude. To achieve a 100% safety margin with 3.5 V, the cut-off pacing threshold is 1.7 V. At discharge it was assessed whether ventricular pulse amplitude remained at < or = 3.5 V or was programmed to > 3.5 V. At pulse generator exchange, pacing threshold was 1.2 +/- 0.5 V, including four patients with pacing thresholds > 1.7 V. Impedance was 587 +/- 189 omega and R wave amplitude was 12.8 +/- 4.8 mV. At discharge, ventricular pulse amplitude remained at the nominal setting in 13 patients, including 2 patients with high pacing thresholds at pacemaker exchange, and was programmed to < 3.5 V in 7 patients. Ventricular pulse amplitude was programmed to > 3.5 V in four patients. Two of these patients had high pacing thresholds (> 1.7 V/0.5 ms) at pulse generator exchange; the other two patients were programmed to > 3.5 V, although 3.5 V already provided a 100% safety margin. After pulse generator exchange, 92% of the patients could be paced with a 3.5-V pulse amplitude. Pacemakers were programmed in four patients > 3.5 V, but in only two of them to obtain a sufficient safety margin. High pacing thresholds at pulse generator exchange did not generally predict high pacing thresholds at discharge.