Poor physical function influences social support among people with Parkinson's disease (PD). Physical function can be improved through exercise, but whether social support influences intervention responsiveness is unclear. This study aimed to (1) examine the influence of 12 weeks of group exercise (tango or walking) on physical function in people with PD who reported low versus high baseline social support, and (2) determine the influence of intervention type on social support effects. This exploratory analysis of an ongoing assessor-blinded randomized controlled trial (1:1) [NCT04122690] included 40 participants with PD (69.24 ± 7.73 years; 26.3% female; Hoehn & Yahr: 2.28 ± 0.58). We used the Multidimensional Scale of Perceived Social Support to categorize participants into high or low support groups. Participants were assessed OFF medication at baseline and 12 weeks using the timed-up-and-go, 360-degree turn, forward, backward, and fast gait speed, 6-min walk, chair stand, and tandem stance tests. Participants were randomized into tango ( = 20) or walking ( = 20) groups and completed 20 sessions within 12 weeks. Left foot tandem stance trended toward improvement regardless of social support ( = 0.06). An interaction among time, support, and intervention indicated that participants in the walking group with low social support improved more on the chair stand test ( = 0.03). No adverse intervention-related events occurred. Overall, high and low social support groups improved similarly following tango or walking interventions. Those in the walking group with low support may have benefited more. Therapeutic intervention targeting physical function regardless of social network in people with PD is important.