This paper draws on ethnographic research in Tanzania to question ideas inherent to mainstream development policy that building social capital can be readily created, used, or substituted for other missing assets, and thereby overcome poverty. The poorest experience clusters of interlocking disadvantage that make it highly unlikely that they can draw on social capital to ameliorate their poverty, or that increased association and participation at community level is necessarily beneficial to them. Moreover, social relationships, collective action, and local institutions may structurally reproduce the exclusion of the poorest. As such, a politically neutral and undersocialized policy focus on strengthening associational life and public participation of the poor is unlikely to lead to their greater inclusion, nor to significant poverty alleviation.