Aims In recent years, non-pulsatile assist devices have been increasingly used for long-term circulatory support. However, it remains to be established whether pulsatile and non-pulsatile systems can ensure a similarly low rate of complications for extended periods of time in elderly patients. Methods and results Of 174 patients implanted between June 1991 and January 2009 [ mean age: 65 +/- 3.7 (60-80) years], 64 were on pulsatile (39 Berlin Heart EXCOR, 18 Novacor, 4 LionHeart, and 3 HeartMate I) and 110 on non-pulsatile assist devices (65 Berlin Heart INCOR, 18 DeBakey, 14 HeartMate II, 7 DuraHeart, and 6 Jarvik2000). In the pulsatile group, the mean support time was 157 +/- 343 (1-1836) days, and in the non-pulsatile group 281 +/- 336 (1-1619) days (P<0.05). The 1-year survival was 15% in the pulsatile group and 36% in the non-pulsatile group, and 2-year survival was 12 and 26%, respectively (P<0.05). In the pulsatile group, 11 patients (17%) were supported for >6 months, 7 patients (11%) for >1 year, and 3 patients (5%) for >2 years, and 3 patients underwent heart transplantation, 3 were weaned, and 3 still have a device. In the non-pulsatile group, 46 patients (42%) were supported for >6 months, 34 (28%) for >1 year, and 14 (13%) for >2 years, and 17 patients are still on support, 9 underwent transplantation, and 3 were weaned. Conclusion Although both types of device can be used for extended periods of time, non-pulsatile systems have a significantly higher survival rate in elderly patients. They allow elderly patients additional years of life in their familiar environment.