BACKGROUND: Abnormal muscle activity during reactive balance may cause balance impairments in Parkinson’s disease (PD) and may be a potential mechanism by which Adapted Tango (AT), an exercise-based balance rehabilitation intervention, improves clinical balance measures. Here, a multidirectional perturbation paradigm was used to quantify how antagonist muscle activity during reactive balance is influenced by 1) PD and impaired balance assessed by standardized behavioral scales and 2) AT. METHODS: Antagonist activation during reactive balance responses to multidirectional support-surface translation perturbations was compared between 1) 31 participants with PD and 13 participants without PD and 2) 30 participants with PD who did (16) or did not (14) participate in AT. Muscle modulation (the ability to activate and inhibit muscles appropriately according to perturbation direction) was quantified using modulation indices (MI, MI180) derived from minimum and maximum EMG activation levels observed across perturbation directions. Modulation was quantified for 100-175 ms (APR1), 70-450 ms (APRX), and 175-250 ms (APRY) after perturbation onset. Clinical measures quantified balance (Berg Balance Scale, BBS; Fullerton Advanced Balance scale, FAB) and gait (Dynamic Gait Index, DGI) performance. RESULTS: In cross-sectional comparisons using MI and APRX, antagonist leg muscle activity was abnormal in participants with PD compared to participants without PD. Linear mixed models identified significant associations between impaired modulation and PD (P<0.05), PD severity (P<0.01), and balance ability (P<0.05), but not age (P=0.10). In the longitudinal examination of AT or Control participants with PD, there was a significant group by time interaction effect on DGI performance, but not on BBS or FAB. Neither the group, time, nor group by time interaction effects were significant for MI in either APRX or APRY. Individual cases showed relationships between FAB and MI changes differing with baseline balance ability. CONCLUSION: This dissertation 1) presents a new method to quantify co-contraction, 2) shows that reduced modulation is associated with PD severity and across PD phenotypes, and with clinical quantifications of balance, and 3) provides evidence suggesting that baseline functional balance ability may be more important to rehabilitation outcomes than age or PD phenotype, with those who stand to benefit most having lower balance ability