The aim of this study was to investigate the effect of battery capacity, internal current drain, and stimulation energy on pulse generators longevity, and if battery impedance measurements can reliably predict pulse generators end-of-life. For this purpose, the records of 577 patients with a mean age of 65 +/- 14 years who had undergone implantation of two different dual chamber pulse generators (PG1: 409; PG2: 168) were retrospectively reviewed. Battery capacity were 2.3 Ah (PG1) and 3.0 Ah (PG2) while current drain at comparable nominal settings was 20 microA (PG1) and 30 microA (PG2) indicating a higher internal current drain of PG2. After a mean follow-up of 46 +/- 23 months, stimulation energy at reprogrammed output settings was significantly higher in PG1 as compared to PG2 (17.1 +/- 0.14) vs 15.5 +/- 0.24 J). Three PG1 (0.7%) and 12 PG2 (7.1%) (P < 0.01) had to be exchanged after a mean of 77.3 +/- 5.3 months (PG1) and 75 +/- 13.5 months (PG2) (P = NS) due to end-of-life being reached. The difference in battery impedances of PG1 and PG2 gained statistical significance 5 years after implantation (1.0 k omega vs 2.4 +/- 6.7 k omega) preceding the significant difference in PG survival after 6 years (98.7 +/- 1.3% vs 90.7 +/- 4.8%). These results indicate that internal current drain is the most important determinant of the pulse generators longevity and that battery impedance can reliably predict end-of-life. Therefore, the essential information about internal current drain should be available for each pacemaker, since it is required for adequate pulse generator selection. Diagnostic functions of dual chamber pulse generators should include measurements of battery impedance.