In this article, we examine participants' talk about qualitative research. We provide empirical support for post structural theorizations of the interview and propose three distinct but related dimensions of qualitative research: emotional, purposive/relational, and epistemic/ontological. In this study, participants often became upset but constructed participation as enjoyable and cathartic. The purpose of participation was to assist the communities to which one belonged. Participation was an active, reflexive practice that reconstructed the self and changed knowledge about one's self. This latter epistemic/ontological dimension of participation appeared to be the most compelling for participants, but it is also the hardest to observe, with implications for how we consider the costs and benefits of participation. We suggest two practical measures for researchers and institutional review boards to consider in light of our findings: routinely asking questions about the research experience in qualitative studies and reformulating patient information statements to particularize them to qualitative research.Keywords: emotions; epistemology; ethics; qualitative research, general; research participation Think back over the qualitative interviews you have conducted and analysed. How often have participants asked about the purpose of a study, questioned your questions, analysed their answers, imagined other participants, or evaluated the potential benefits arising from a project? Such a comment or question is a kind of meta research moment, an instance in which the usual interview procedure-the researcher asking questions and the participant answering-is overturned, drawing attention to the interaction itself. You might remember several meta research moments from your own career, particularly if they involved a potential loss of face for you as interviewer or produced a key insight. You have probably forgotten many others, as these moments are usually treated as peripheral. The synthesis of raw data into powerful ideas and recommendations is difficult. In analysis we strive for focus, for explanatory power, for coherent theorization, and for description layered thickly rather than spread thin so that we can make a convincing case to our readership and enhance the usefulness of our work. In analysis we create succinct publications and presentations by prioritizing and synthesizing data. The meta moments referred to above are usually deprioritized,