BACKGROUND: Little is known about outcome and settings adaptations after replacement of constant-voltage non-rechargeable implantable pulse generator (CV-nrIPG) by constant-current rechargeable IPG (CC-rIPG). OBJECTIVE: To determine the feasibility and safety of replacing a CV-nrIPG by a CC-rIPG in Parkinson's disease (PD) and the subsequent outcome. METHODS: A prospective cohort of thirty PD patients, whose CV-nrIPG was replaced by a CC-rIPG in University Hospital of Lyon between January 2017 and December 2018 (rIPG group) and 39 PD patients, who underwent the replacement of a CV-nrIPG by the same device in 2016 (nrIPG group), were enrolled in this study. Three surgeons performed the operations. Duration of hospitalization for the replacement as well as the number of in or outpatient visits during the first 3 months after the surgery were recorded. In the rIPG group, we compared preoperative DBS settings and the theoretical amplitude estimated using Ohm's law to the amplitude used at the end of follow-up. We assessed patients' and clinicians' opinion on the patient global functioning after the replacement using Clinical Global Impression score. RESULTS: Duration of hospitalization (P=0.47) and need for additional hospitalizations (P=0.73) or consultations (P=0.71) to adapt DBS parameters did not differ between the two groups. Neurological condition (CGI score) was considered as unchanged by both patients and neurologists. Final amplitude of stimulation using CC-rIPG was not predicted by Ohm's law in most cases. CONCLUSIONS: Replacing CV-nrIPG by CC-rIPG is safe and well tolerated but require neurological expertise to set the new parameters of stimulation.