Bias in self-reports is one of the key problems which confronts researchers interested in using qualitative research techniques to understand individuals’ cognitive and conceptual models of health behaviours. This is a problem that is especially apparent, for example, in research around addictive behaviours, where the fundamental trustworthiness of respondents’ accounts may be in doubt. This research note discusses the problems associated with using conventional qualitative methods in health research and suggests a means of avoiding these.