This article begins by raising issues around the way in which ethical approval for research is managed in university settings, where committees often base their assumptions on a principlist approach making a number of assumptions that we consider to be contestable, such as a neat separation between researcher and researched. However, collaborative action research, we argue, takes issue with the "objectification" of research participants. It often blurs the distinction between participant and researcher, particularly when an element of self-study is included. Moreover, the collaborative nature of action research problematises the question of who is researcher and who is researched, raising issues around anonymity, the "ownership" of findings and dissemination. In response to some of these issues, we have developed a set of eight principles we derive from our "version" of collaborative action research and apply them in a discussion of a number of case studies from our own setting, where researchers have faced a number of dilemmas in attempting to work within the terms of reference imposed by conventional university-based ethical approval procedures. In conclusion, this article indicates some implications for university-based action researchers and makes recommendations about the forms of ethical scrutiny within the university that would be most appropriate and searching for collaborative action-based enquiry. (Contains 1 table and 2 notes.)