This study aimed to examine the influence of physical and cognitive leisure activities on the survival of older adults in China, while also exploring the potential mediating and moderating effects of general health. This study utilized the China Longitudinal Healthy Longevity Survey (CLHLS) datasets spanning from 2008 to 2018, and 10,347 eligible participants were included. The primary study outcome was all-cause mortality, and independent variables included physical leisure activities (PLA), cognitive leisure activities (CLA), and self-rated general health. Three sets of covariates were adjusted, including socio-demographic characteristics, health behaviors, and health status. The longest survival time was the older people participating in PLA & CLA (mean = 50.31 months), while those participating in neither exhibited the lowest (mean = 29.60 months). Significant differences in survival status were observed in different types of leisure activities participation (Log-rank test, Chi-square = 576.80, p < 0.001). Cox regression indicated that PLA (HR = 0.705, 95% CI: 0.651-0.764), CLA (HR = 0.872, 95% CI: 0.816-0.933), and the both PLA & CLA (HR = 0.700, 95% CI: 0.656-0.747) were protective factors for the survival. Additionally, general health significantly moderated the relationship between PLA and reduced mortality risk (Coefficient = -0.089, p = 0.042). While CLA indirectly influenced the survival through general health (Coefficient = -0.023, p < 0.001). For the older people participating in PLA and CLA, general health played mediating (Coefficient = -0.031, p < 0.001) and moderating (Coefficient = -0.026, p = 0.013) role in the relationship between leisure activities and survival. Leisure activities and self-rated general health were important predictors of survival of the older adults, and general health exhibited a mediator and moderator in the relationship between leisure activities and survival status.