The aims of this thesis were two-fold: to examine the phenomenology of auditory verbal hallucinations (A VH) and to compare the experiences of voice-hearers who have, and do not have, a need for care. The theoretical foundations of the thesis lie in psychological models of psychosis which suggest that voices are not essentially pathological and that distress results from beliefs about voices that are influenced by life history. F01ty voice-hearers were recruited, half of whom had a psychiatric diagnosis and had received treatment for their voices and half of whom who had no need for care in relation to their voices and no psychiatric history. All participants completed interview and questionnaire measures of their experiences of voices, beliefs about voices and life history. Ten voice-hearers (three of whom were clinical participants) were scanned using functional magnetic resonance imaging whilst they were hearing voices. The results of the three studies were largely consistent with each other and previous research and theory. The two groups' voices were similar in topography and biological basis but the clinical group's voices were more negative in content and caused more distress and disruption. There were significant group differences on a number of cognitive and environmental variables; the clinical group had more negative beliefs about voices, thoughts and other people as well as more negative life histories involving interpersonal trauma and relationship difficulties. Mediational analyses suggested that adverse life experiences (trauma and insecure attachment) influence the distress associated with voices, through their influence on beliefs about voices. The thesis concludes with a discussion of the current findings and the consequent implications for future research on A VH and their treatment.