Parkinson's Disease (PD) is a progressive neurodegenerative disorder that affects movement and balance control. Falls are a common and often debilitating consequence of PD, and reactive balance control is critical in preventing falls. This dissertation aimed to determine the adaptability and neural control of reactive balance responses in people with PD. Aim 1 investigated whether people with PD at risk for falls can improve their reactive balance responses through a 2-week, 6-session training protocol. The study found that reactive step training resulted in immediate and retained improvements in stepping, as measured by the anterior-posterior margin of stability (MOS), step length, and step latency during backward stepping. The second aim explored the neural mechanisms behind eliciting and learning reactive balance responses in PD. The study investigated the white matter (WM) correlates of reactive stepping and responsiveness to step training in PD. White matter was not significantly correlated with any baseline stepping outcomes. However, greater retention of step length was associated with increased fractional anisotropy (FA) within the left anterior corona radiata, left posterior thalamic radiation, and right and left superior longitudinal fasciculi. Lower radial diffusivity (RD) within the left posterior and anterior corona radiata were associated with retention of step latency improvements. These findings highlight the importance of WM microstructural integrity in motor learning and retention processes in PD. The third aim examined the role of the somatosensory system in reactive balance control in people with PD. The tactile and proprioceptive systems were perturbed using vibrotactile stimulation during backward feet-in-place balance responses. The results showed that tactile and proprioceptive stimulation had minimal impact on reactive balance responses. Small effects were observed for delayed tibialis anterior (TA) onsets with proprioceptive stimulation at a medium intensity. Overall, this dissertation provides insights into improving reactive balance responses and the underlying neural mechanisms in PD, which can potentially inform the development of targeted interventions to reduce falls in people with PD.