Aim To investigate whether being part of a community organization interfered with older adults’ overall functionality. The magnitude of responses to a community exercise program based on functional circuits in socially active and socially non‐active older adults was also investigated. Methods A total of 200 older adults aged ≥60 years from Requinoa, Chile, participated in the study. Participants were separated into two groups according to the level of social participation: socially active (SA) and socially non‐active (SNA). During an evaluation, data regarding the presence of comorbidities (Charlson Comorbidity Index), disability level (Modified Health Assessment Questionnaire), dynamic balance (Timed Up and Go test), muscle force (handgrip dynamometry) and sociodemographic conditions were collected. Participants followed an exercise program consisting of a functional exercise circuit including balance, resistance and aerobic exercises, twice a week, for 12 weeks. Results Both the Charlson Comorbidity Index and disability were higher in SNA compared with SA participants on inclusion. Both groups improved disability (Δ–0.25 patients [−0.5 – −0.0625 patients] for SNA and (Δ–0.125 patients [−0.5–0 patients] for SA) and balance (Δ–2 s [−3–0 s] for SNA and (Δ–1 s [−3–0 s] for SA) after training (P < 0.05 for all). Changes from baseline were statistically higher in the SNA group. Handgrip force improved only in participants in the SNA group (Δ1.7 kg [0.6–2.8 kg], P = 0.0001). No differences, however, were observed between the magnitude of improvements of handgrip between groups. Conclusions Although socially non‐active older adults had more comorbidity and disability than their socially active counterparts, they showed a higher response to a community exercise program. Geriatr Gerontol Int 2017; ••: ••–••.