Alexander Leighton's seminal work has clearly demonstrated how ethnographic experience provides the rich cultural context in which epidemiological data are best interpreted. This article reviews recent trends in cultural epidemiology, and especially the emergence of the EMIC (Explanatory Model Interview Catalogue) as a quantitatively oriented tool designed to assess culture. It is suggested that such efforts do not reflect more recent trends in culture theory, and tend to view 'cultures' as easily bounded and largely homogenous units to facilitate the generation of quantitative data. It is argued that cultural epidemiologists should take a step back and ask, 'what is the culture in question here?' and 'how do I know if it is appropriate to place any given member of my sample into a specific cultural category?' before proceeding with any 'culturally appropriate' instrument. The answer to these questions begins with a return to ethnography as a means to elucidate and describe culture within the context in which it is being presented and studied.