Background: Concept maps have been used extensively for developing higher order thinking skills and are considered significant artefacts in constructing understanding in educational contexts. Increasingly, they are being used as a tool to chart a way towards 'new understanding' rather than recording 'accepted knowledge'. This study is set in an academic development department in a UK higher education institution in which previous research projects have utilised concept map-mediated interviews as a tool in data collection. This paper reports on the relationship between the process of the concept map-mediated interview and the resulting concept map and focuses on the talk during the interview process. Purpose: The purpose of the study was to explore the co-constructed nature of the concept map which resulted from the concept map interview. The research question was: how is the concept map accomplished through and in the interview talk?Sample: The three researchers and authors of this paper are colleagues in an Academic Development department in a UK higher education institution. The focus of the interview was to probe the research perspective underpinning the practice of one of the authors. Design and methods: The study used a qualitative, unstructured concept map interview. The aim of the interview was to elicit an understanding of one of the authors' research frame and how it influenced her work with staff. The interviewer noted labels on post-it notes during the interview which both participants then arranged on a sheet of paper. The interview lasted 36 minutes and was transcribed verbatim. Sociocultural discourse analysis was used to examine the trajectory of concepts in the interview talk. Results: The results highlight the collaborative nature of the interview and how the concept map is co-constructed through the interview talk. We demonstrate how the concept map is co-constructed "through" and "in" the dialogue between interviewer and interviewee, not "as a result of" the interview. Results also reveal how the context of acquaintance interviews impacts on the co-construction and thus the resulting concept map. Conclusions: A concept map which results from such an interview is co-constructed with the interviewer playing a pivotal role in the talk and the mapping. The implications are that the interview as research tool needs to be recognised as a site for the co-construction of ideas and perspectives. Concept maps resulting from interviews need to be recognised as co-constructed. A further implication for research methods is that the transcripts from the interview itself can be used as data to provide a richer understanding of the concept map.