Lifestyle activity engagement is a modifiable factor for cognitive decline. We aimed to identify lifestyle patterns (LPs) among community-dwelling older adults in the pre-dementia stages and to explore the links between LPs, cognitive function, and individual characteristics. 702 older Chinese adults were recruited. Three LPs were identified by latent class analysis: active aging lifestyle pattern (AALP), leisure lifestyle pattern (LLP), and work-centered lifestyle pattern (WLP). AALP refers to participation in various activities that are meaningful to individuals and benefit their well-being. LLP is the pattern of activities aimed at recreation. WLP refers to the LP where individuals are most likely to engage in work-related activities. However, only AALP is protected against mild cognitive impairment (MCI). Multinomial logistic regression models revealed the differences in individual characteristics among participants with different LPs, indicating the importance of tailored intervention strategies. As a protective factor against MCI, AALP should be highlighted in community-based care.