In 55 consecutive patients with the same dual chamber pacemaker (Relay, Intermedics) and different pacing leads, the influence of different safety margins for pacing on battery current was investigated. 2.8 +/- 0.9 years after implantation, atrial and ventricular pulse-width thresholds (tRS) (ms) were determined at 0.5, 1.0, and 2.0 V, and the charge delivered at threshold was telemetered. If tRS was < 1.50 ms at 0.5 V, an amplitude of 1.0 V was programmed in the atrium and the ventricule; if tRS was < 1.50 ms at 1.0 V, then an amplitude of 2.0 V was chosen. Two times the charge threshold (2 x QRS), two times the voltage threshold (2 x URS), and three times the pulse-width threshold (3 x tRS) were programmed as the safety margins for pacing. With every safety margin, battery current (IBat) (microA) was averaged from 5 telemetric readings in D00 mode with 70 bpm. IBat was significantly lower with 2 x QRS as compared with 2 x URS (13.43 +/- 1.0 vs. 14.20 +/- 1.2 microA, p < 0.01) and as compared with 3 x tRS (13.99 +/- 1.2 microA, p < 0.05). Pacemaker longevity derived from these current data was significantly longer with 2 x QRS (112 +/- 8 months) as compared with 2 x URS (106 +/- 9, p < 0.01) and as compared with 3 x tRS (108 +/- 8, p < 0.05). If current consumption is compared intraindividually in dependance on the programmed amplitude, battery current is significantly lower at 1.0 V as compared with 2.0 V resulting in a mean reduction of 0.63 microA (-4.9%, p < 0.05) and an average gain in longevity of 5 months. This applies to every safety margin tested. Differences in battery current caused by the safety margins will translate into a greater gain in longevity in future pacemaker models with reduced internal current consumption.