This paper argues that emerging 'post-medical geographies' require attention to the methodological in order to fully appreciate how different geographical knowledges are produced and contextualized within the politics of research relationships. 'Geographies of mental health and illness' are focused upon in order to argue that the 'peopling' of health research should also be accompanied by debate about what sorts of methodologies we employ in accessing these minds/bodies and voices. The research interview is a primary focus here. A critique of psychoanalytic approaches to geographical research argues that such 'models' of interpretation and management can mean that participants or research 'subjects' can be framed in almost diagnostic categories of behaviour. Empirical examples of mental health research in Nottingham are used to argue that more flexible approaches which pay attention to perceived dualisms (such as 'sanity' and 'insanity'), negotiation, embodiment, socio-spatial contexts and content within the interview situation may aid in understanding the politics which encompass geographical health research.