Purpose: The aim of this study is to examine the reciprocal causal relationship between social activities and health with reference to the cognitive function level among community-dwelling older adults. Methods: We conducted a longitudinal data analysis using the 4^th (Time 1) and 5^th (Time 2) waves from the Korean Longitudinal Study of Ageing adopting cross-lagged panel analysis (CLPA). Results: A total of 3,473 community-dwelling older adults were included in the analysis: 2,053 in the normal cognitive function group, 912 in the mild cognitive impairment group, and 508 in the moderate to severe cognitive impairment group. The CLPA revealed that higher levels of social activities at Time 1 significantly influenced better subjective health perception at Time 2 in all three groups. In addition, better subjective health perception at Time 1 significantly influenced higher levels of social activities at time 2 only in the normal cognitive function group. Conclusion: This study demonstrates that participating in social activities has positive effects on health regardless of the cognitive function level in community-dwelling older adults. This finding suggests that there is a need to develop strategies that can be used to encourage older adults with cognitive decline to maintain participation in social activities.