Social capital has been recently held up as a conceptual framework to build a bridge between the diverse disciplines involved in rural development. However, despite its potential and the impressively rapid take-up of the concept by the community of development professionals, it remains an elusive construct. No definition is yet generally accepted and many definitions are in use. Recently, social capital in the form of social networks has gained much attention in rural development theory and empirical research. But social networks or structural components of social capital are a largely missing dimension in income and poverty analysis. Moreover, most research on social capital assumes that it is a uniform entity. Therefore, the effects of different forms of social capital on household outcome are rarely investigated. The objective of this discussion paper is to make the concept of social capital more tangible for empirical research in the area of rural development and to bring more structure into the conceptual framework of social capital. On the basis of an extensive literature review, this work proposes a lean and clear definition of social capital: Social capital is conceived as networks plus resources, (e.g. credit, information). Moreover, social capital is assumed to be not a homogeneous entity. Hence, it is necessary to distinguish different forms of social capital. For analytical purposes, the separation into so-called bonding and bridging capital seems to be most appealing. These two forms of social capital can be operationalized as function of an agent’s so-called weak ties (e.g. acquaintances) and so-called strong ties (e.g. close relatives).