Although ample evidence suggests that high-frequency deep brain stimulation (DBS) is an effective therapy in patients with Tourette syndrome (TS), its pathophysiology and the neurophysiological mechanisms underlying these benefits remain unclear. The DBS targets mainly used to date in TS are located within the basal ganglia-thalamo-cortical circuit compromised in this syndrome: the medial and ventral thalamic nuclei, which are way stations within the circuit, the globus pallidus and the nucleus accumbens. Neuronal activity can be electrophysiologically recorded from deep brain structures during DBS surgery (intraoperative microrecordings) or within few days after DBS electrode implantation (local field potentials, LFPs). Recordings from the thalamus in patients with TS showed that the power in low-frequency oscillations (2–15 Hz) was higher than power in high frequency oscillations (<45 Hz) and that activity in gamma band (25–45 Hz) increases when patients’ clinical status improved. Effective thalamic DBS for tic reduction seems to increase high frequency band oscillations (25–45 Hz). The same oscillatory pattern persists after DBS for 1 year, therefore showing that in TS DBS does not induce persistent neuroplastic changes in the neural activity in the stimulated structures. Neurophysiological recordings from deep brain structures suggest that tics originate not from the cortex but from neuronal dysfunction in deep brain structures such as the thalamus and globus pallidus. In conclusion, DBS can induce its beneficial effects in TS by modulating specific neural rhythms in the cortico-basal ganglia thalamic network. DBS could reduce tics related increased low-frequency activity by shifting the basal ganglia-thalamic oscillation power to higher frequencies.