Abstract
There is increasing interest in measuring ‘social capital’ to explore the ways social relations contribute to complex health inequalities between population groups. Most research has relied on the health survey to explore social capital concepts but this research methodology is problematic. This article examines two different theoretical perspectives that can inform the measurement of social capital and describes the different research methodologies associated with each. One perspective uses the answers individuals give to social survey questions; the other relies on more complex historical and ethnographic methods. Examples of research using both methodologies are described. The paper argues that key concepts like ‘social relations’ and ‘social capital’ are not adequately measured using the individual answers that people report on surveys and that much work still needs to be done to validate what answers to survey questions actually represent.