Aim We tested the mediating effect of decline in social participation on urinary incontinence (UI) and negative mood in older adults by sex and marital status. Methods We carried out secondary analysis of data collected from 5301 Chinese adults aged 60 years or older in Hong Kong who had completed an initial screening instrument for subsidized long‐term care services in 2010. Path analysis within structural equation modeling was carried out. Results Satisfactory model fit was obtained: male–married group (n = 1949, standardized root means squared residual [SRMR] = 0.034, robust root mean square error of approximation [R‐RMSEA] = 0.045, robust comparative fit index [R‐CFI] = 0.965), male–other group (n = 519, SRMR = 0.023, R‐RMSEA = 0.011, R‐CFI = 0.988), female–married group (n = 948, SRMR = 0.018, R‐RMSEA = 0.002, R‐CFI = 1.000) and female–other group (n = 2251, SRMR = 0.023, R‐RMSEA = 0.048, R‐CFI = 0.970). In the male–married subsample, UI had both a significant direct effect (β = 0.046) and significant indirect effect on mood through decline in social participation (β = 0.021); for the female–other subsample, UI (β = 0.058) and decline in social participation (β = 0.335) had significant direct effects on negative mood. Decline in social participation had a significant direct effect on negative mood in the male–other subsample (β = 0.306) and the female–married subsample (β = 0.325). Conclusions Decline in social participation mediated the relationship between UI and negative mood in married older men, but not male–other group or women. It is important to recognize these sex and marital status differences, and expand social participation opportunities for UI patients. Geriatr Gerontol Int 2017; ••: ••–••.