This article concerns the manner in which group interaction during focus groups impacted upon the data generated in a study of adolescent sexual health. Twenty-nine group interviews were conducted with secondary school pupils in Ireland, and data were subjected to a qualitative analysis. In exploring the relationship between method and theory generation, we begin by focusing on the ethnographic potential within group interviews. We propose that at times during the interviews, episodes of acting-out, or presenting a particular image in the presence of others, can be highly revealing in attempting to understand the normative rules embedded in the culture from which participants are drawn. However, we highlight a specific problem with distinguishing which parts of the group interview are a valid representation of group processes and which parts accurately reflect individuals' retrospective experiences of reality. We also note that at various points in the interview, focus groups have the potential to reveal participants' vulnerabilities. In addition, group members themselves can challenge one another on how aspects of their sub-culture are represented within the focus group, in a way that is normally beyond reach within individual interviews. The formation and composition of focus groups, particularly through the clustering of like-minded individuals, can affect the dominant views being expressed within specific groups. While focus groups have been noted to have an educational and transformative potential, we caution that they may also be a source of inaccurate information, placing participants at risk. Finally, the opportunities that focus groups offer in enabling researchers to cross-check the trustworthiness of data using a post-interview questionnaire are considered. We conclude by arguing that although far from flawless, focus groups are a valuable method for gathering data about health issues.