Identifying neural activity biomarkers of brain disease is essential to provide objective estimates of disease burden, obtain reliable feedback regarding therapeutic efficacy, and potentially to serve as a source of control for closed-loop neuromodulation. In Parkinson's Disease (PD), microelectrode recordings (MER) are routinely performed in the basal ganglia to guide electrode implantation for deep brain stimulation (DBS). While pathologically-excessive oscillatory activity has been observed and linked to PD motor dysfunction broadly, the extent to which these signals provide quantitative information about disease expression and fluctuations, particularly at short timescales, is unknown. Furthermore, the degree to which informative signal features are similar or different across patients has not been rigorously investigated. Here, we recorded neural activity from the subthalamic nucleus (STN) of patients with PD undergoing awake DBS surgery while they performed an objective, continuous behavioral assessment. This approach leveraged natural motor performance variations as a basis to identify corresponding neurophysiological biomarkers. Using machine learning techniques, we show it was possible to use neural signals from the STN to decode the level of motor impairment at short timescales (as short as one second). Spectral power across a wide range of frequencies, beyond the classic "beta" oscillations, contributed to this decoding. While signals providing significant information about the quality of motor performance were found throughout the STN, the most informative signals tended to arise from locations in or near the dorsolateral, sensorimotor portion. Importantly, the informative patterns of neural oscillations were not fully generalizable across subjects, suggesting a patient-specific approach will be critical for optimal disease tracking and closed-loop neuromodulation.