Excessive beta oscillations in the subthalamic nucleus are established as a primary electrophysiological biomarker for motor impairment in Parkinson’s disease and are currently used as feedback signals in adaptive deep brain stimulation systems. However, there is still a need for optimization of stimulation parameters and the identification of optimal biomarkers that can accommodate varying patient conditions, such as ON and OFF levodopa medication. The precise boundaries of “pathological” oscillatory ranges, associated with different aspects of motor impairment, are still not fully clarified. In this study, we hypothesized that analyzing periodic and aperiodic components of subthalamic nucleus activity separately and identifying functionally distinct subranges within 8-35 Hz based on oscillatory properties may reveal robust biomarkers for specific aspects of motor impairment. We analyzed subthalamic nucleus activity of 14 patients with Parkinson’s disease. Local field potentials were recorded at rest from externalized electrodes postoperatively, both before and after levodopa administration. We showed that levodopa administration suppressed oscillations across a broad frequency range (11-32 Hz) and increased the slope of the aperiodic component. Changes in the aperiodic slope correlated with motor symptom alleviation. Periodic activity was linked to motor symptom severity: peak amplitude within the 14-20 Hz range correlated with overall motor impairment in the OFF state, while the 7-11 Hz range was associated with bradykinesia in the ON state. Our findings suggest that, in addition to low beta, alpha oscillations and the aperiodic component may serve as promising biomarkers for motor impairment and potential feedback signals in adaptive DBS systems.