This paper considers the health implications of access to social capital (SC) using data from a survey of households in two suburbs in Adelaide, Australia. SC was conceptualised as comprising 'infrastructure'--the networks and values that facilitate access to resources; and 'resources'--the resources available through this infrastructure. Questionnaires were delivered to all households in the area, asking the adult with the next birthday to complete it. In all, 530 (50%) were returned. A partial least-squares path analysis was undertaken using demographic, SC and health latent variables, and a measure of perceived relative advantage. Three infrastructure (values, formal networks and informal networks) and four resource (help, acceptance by neighbours, civic activities and feelings of control) variables were considered. Mental and physical health were measured using the SF-12. The values variable was associated with all the resource variables, the informal networks variable was related to help, and the formal networks variable was associated with civic actions. There were significant sociodemographic differences in a number of the infrastructure and resource variables, as well as mental and physical health. Those who were better off materially also had greater access to elements of SC, and reported better health. Values, informal networks, help, and control were all directly or indirectly positively associated with better mental health. No SC variables were associated with physical health. Perceived relative advantage was positively associated with a number of SC variables and also mental and physical health. The implications for health promotion are discussed.