Background. There is debate about undertaking sensitive research with vulnerable populations. Primarily, the literature has focused on informed consent, confidentiality and the principle of beneficence, with little discussion about data collection methods. Aim. This paper discusses the challenges of conducting unstructured interviews when participants have a major depressive illness. Issues arose during a phenomenological study that explored the meaning of being nurtured with seven people who were hospitalized for depression. Findings. The depressive illness and treatment were found to impact on participants' articulation and recalling of their experience, and raised ethical concerns about their informed consent. Personal engagement with participants raised the ethical issue of research vs. therapy. Furthermore, participants being in a hospital complicated the necessity for privacy. The methodological issue of bracketing of assumptions was deemed to be important to 'see' the phenomenological relevance of patients' experiences. Conclusion. Knowledge and experience are required when conducting unstructured interviews. Debates about the challenges of unstructured interviewing needs to be highlighted in research texts to assist novice researchers. Support from an experienced research mentor would assist novice interviewers through the interview process and provide post-interview debriefing.