Background Few studies have examined the potential difference in the relationship between bonding versus bridging social capital and health outcomes. We sought to examine the association between these different types of social capital and self-rated health in a population-based study.

Methods In February 2009, 4000 residents of Okayama City (aged 20–80 y) were randomly selected for a survey on social capital and health. The survey asked about participation in six different types of associations: Parents and Teachers Association, sports clubs, alumni associations, political campaign clubs, citizen's groups and community associations. We distinguished between bonding and bridging social capital by asking participants about their perceived homogeneity (with respect to gender, age and occupation) of the groups they belonged to. ORs and 95% CIs for poor health were calculated.

Results Bridging social capital (ie, participation in groups involving people from a diversity of backgrounds) was inversely associated with poor health in both sexes and women appeared to benefit more than men. Compared to those who reported zero participation, high bridging social capital was associated with a reduced odds of poor health (OR 0.25, 95% CI 0.11 to 0.55) in women after controlling for demographic variables, socioeconomic status, smoking habit and overweight. By contrast, bonding social capital was not consistently associated with better health in either gender.

Conclusions The present study suggests that bonding and bridging social capital have differential associations with health and that the two forms of social capital need to be distinguished in considering interventions to promote health.