That organizational involvement has a positive impact on political action is a well-established finding in empirical research around the world. To account for this, theorists since Tocqueville have pointed to the returns in human capital, in particular ‘civic skills’, yielded by associations. This article, by contrast, is a study of whether social capital theory can help explain the same effect. According to the logic of ‘weak ties’, organizational involvement provides bridging social capital by connecting the individual to a wider range of people. As a result, the input of requests for participation increases and this ultimately leads to more activity. Unspecified in this argument, however, is what aspect of associational memberships is most conducive to such weak ties: the sheer number of memberships, or the extent to which one's memberships provide links to people of dissimilar social origin. In an unprecedented empirical test based on survey data from Sweden in 1997, it is shown that being connected to multiple voluntary associations is what matters for political activity, not the extent to which one's memberships cut across social cleavages. Moreover, the social capital mechanism of recruitment is more important in explaining this effect than the human capital mechanism of civic skills, since the former can account for why even passive members, not just organizational activists, may become more prone to take political action.