The links between social capital and mental wellbeing are established but the direction of the social capital-wellbeing relationship is rarely systematically examined. This omission undermines the validity of social capital as a basis for health interventions. The aim of this paper was to explore the short-term (one-year) reciprocal relationship between community participation – an important component of social capital – and mental wellbeing. We used nationally representative Household, Income and Labour Dynamics in Australia (HILDA) Survey data, 2005–11. The HILDA Survey is an annual cohort study from which was extracted a sub-sample of panel data (the same people participating across multiple waves) enabling us to use fixed effects regression methods to model the longitudinal association of mental health and participation controlling for individual heterogeneity. The results showed that better mental wellbeing in one year was generally related to more community participation the next year, while greater past community participation was linked to better mental wellbeing the next year independent of (i) initial mental wellbeing, (ii) multiple potentially confounding factors and (iii) unobserved and time-constant heterogeneity. Political participation was marginally related to worse mental health in both directions. The results also showed that the association between community participation and mental wellbeing the next year is weaker for those with poor initial wellbeing than for initially healthier respondents. Our findings may inform the trial and scientific evaluation of programs aimed at increasing informal social connectedness and civic engagement to promote mental wellbeing.