SUMMARY Auditory verbal hallucinations (AVH) afflict more than 80% of schizophrenia (SZ) patients and are treatment resistant up to 40-45%. AVHs increase the risk of suicidal, aggressive behavior, reduced work attainment and impairment in specific cognitive domains. Considering that 25-50% of persons with SZ are treatment-resistant (TR-SZ), there is unmet public health need to treat persistent AVH in TR-SZ. SZ is a disorder with disruptions within corticostriatothalamic circuits ((CST) thus potentially amenable to modulation via Deep Brain Stimulation (DBS). DBS-SZ may hypothetically modulate AVH through its effect on projections from superior temporal gyrus (STG) to basal ganglia (BG), via the substantia nigra pars reticulata (SNr) and mediodorsal nucleus of the thalamus (MDN). There is evidence that the SNr-MDN-STG loop is dysfunctional in SZ and lesions within this loop cause new onset SZ-like hallucinations. Modulation of the SNr-MDN-STG loop could result in treatment of persistent hallucinations in SZ. In support of this hypothesis, we have preliminary evidence that bilateral SNr DBS induces sustained remission of chronic AVH in a TR-SZ patient and their improvement in a second patient. We hypothesize (Aim 1) that SNr DBS decreases AVH measured by in-clinic BPRS, and at-home-Ecological momentary assessment technique by at least 20% without serious adverse events when comparing baseline to 60 weeks DBS stimulation. For more granular examination of DBS effect on AVH, we will use in-clinic Auditory Vocal Hallucinations Rating Scale (AVHRS), and the Scale for the Assessment of Positive Symptoms (SAPS) (Aim 1). A potential neural mechanism of SNr DBS modulation for AVH might be restoring oscillatory activity in the SNr-MDN-STG loop. Abnormalities in γ oscillations (30–100 Hz) of the electroencephalogram are ubiquitous in SZ. SNr recordings in persons with SZ is uncharted, but low-γ oscillations have been recorded from the SNr in ketamine- treated rats, a well-established SZ model--these are hypothesized to reflect cognitive and sensory, rather than motor impairment. We hypothesize that AVHs in SZ are associated with excessive gamma power in the SNr, and that these abnormal SNr gamma oscillations can be targeted with bilateral SNr DBS resulting in reduced AVH. We will collect home LFP recordings to assess changes in SNr LFP gamma power in DBS on vs off and explore if DBS-related oscillatory changes correlate with AVH changes (Aim 2). Since AVH are associated with impairments in saccadic eye-movements s, we will also explore if DBS-related oscillatory changes correlate with saccadic impairments and their improvements Aim 2.)