Parkinson’s Disease (PD) is a neurological disease that affects motor function. Symptoms include muscle rigidity, tremors, slowed movements, and altered voice. Deep Brain Stimulation (DBS) is a therapeutic intervention addressing PD symptoms, by implanting an electrode into the brain, delivering electrical impulses to areas affected by PD. DBS patients, however, must come into clinic every few months for multi-hour reprogramming sessions to adjust DBS settings according to the progressions of their symptoms, making DBS treatment an arduous and expensive process. This significantly limits the accessibility of DBS, because only limited locations/providers who can offer this service. Our lab has developed a computer algorithm that can derive digital biomarkers, indicators for the severity/presence of a disease, for PD from voice samples. We aim to investigate the correlation between patient voice and physical symptoms, with manipulation of stimulation in the on/off states. Our findings could aid clinicians in their monitoring, management and adjustment of DBS protocol, without the need for patients to come into clinic. We tested this by obtaining motor scores via the Unified Parkinson’s Disease Rating Scale (UPDRS), and gathering voice samples with DBS turned on and off, during patients’ visit at UW clinics. Voice samples were analyzed for specific biomarkers, using our machine learning algorithm. We expected to observe different voice-associated biomarkers present with DBS turned on/off, to demonstrate a correlation between DBS and voice/motor symptoms. Once the relationship between DBS and voice is established, the ultimate goal is to develop a closed looped DBS device which can auto tune itself based on patient voice alone. In an age where audio collection devices, such as smartphones are so accessible, the use of voice data has exciting potential as a clinical tool, optimizing DBS therapy protocol and efficiency for administering clinicians and patients with PD.