People living with chronic illness experience impairment in various ways, not the least of which is how they are sometimes marginalised by the people with whom they interact. Over the last few decades, as social science research has moved away from the biomedical model, research methodologies have been developed to allow the voices of people with illness or disabilities to be heard and not only to be represented by others. However, these methodologies may not go as far in redressing power imbalances as was hoped, and participants' voices are often still mediated and subjugated to the researcher's requirements. As a person who has lived with a chronic condition all my life, I am concerned about how I am heard and by whom, as this has often affected my self-perception and sometimes even my safety. I am also concerned about doing violence to those about whom I speak and disempowering them further. My doctoral research concerned the problematic of writing about my own experience of a chronic illness. I chose a methodology, autoethnography, that allowed me to write solely about my own experience. In so doing, I was able to consider the complexity of my own academic and narrative voices, individually and in combination. In this paper I explore the methodological and epistemological concerns around my decision to use autoethnography, as well as the sometimes surprising issues I navigated when doing so. One of these issues was the juxtapositioning of different types of texts I had written. In isolation, some of the texts show a clear influence of the very discourses to which I was trying to provide counter-narratives. When read together with other texts, they reveal a complex web of paradoxes, tensions, and silences, which allowed me to generate new narratives and to question assumptions -my own and other people's.