Background and Purpose:  The aim of the study was to analyse the lifetime of Soletra implantable pulse generators (IPG) in deep brain stimulation (DBS) of the globus pallidus internus (GPi) for dystonia, depending on stimulation parameters and the total electrical energy delivered (TEED) by the IPG. Methods:  In a prospective series of 20 patients with GPi DBS for dystonia, we recorded IPG longevity and stimulation parameters over time. An evaluation of the TEED was performed using the previously suggested equation (voltage2 × pulse width × frequency)/impedance  × 1 s. Results:  During median follow‐up of 57 months (range 23–79 months), 64 IPGs were replaced because of battery depletion or end of life signal. We found a mean IPG longevity of 25.1 ± 10.1 (range 16–60) months, which was inversely correlated with the TEED (r = −0.72; P < 0.001). IPG longevity was not different between bipolar and monopolar stimulation (24.9 ± 10.8 vs. 25.4 ± 9.0 months, P = 0.76). Incongruously, the mean TEED applied throughout the lifetime cycle was significantly higher in patients with bipolar compared with monopolar stimulation (584 ± 213 vs. 387 ± 121 Joule; P < 0.01). Conclusions:  Battery lifetime in GPi DBS for dystonia is substantially shorter compared with that reported in DBS for Parkinson’s disease, caused by a considerably higher voltage and greater pulse width and therefore a higher TEED applied during the battery lifetime cycle. The commonly used equation to calculate TEED, however, seems to be correct only for monopolar, but not bipolar stimulation.