This thesis forms part of the examination for the degree of Doctor of Clinical Psychology (ClinPsyD), in the Faculty of Biological Medicine and Health at the University of Manchester. The thesis has been written by Lucy Shattock and submitted in July 2016 for examination in September 2016. Paper 1 provides a systematic review of studies investigating therapeutic alliance in the context of psychological interventions for people with psychosis. Twenty-six studies were identified that satisfied the inclusion criteria. The review comments on the nature of alliance across studies, the extent to which alliance predicted outcome, as well as client, therapist and therapy-related factors associated with alliance. Clinical implications are discussed and research recommendations are made. Paper 2 details an investigation into the relationships between attachment, therapeutic alliance and service engagement in a Black sample with psychosis. Twenty eight participants with non-affective psychosis were recruited across a number of outpatient and inpatient services. Participants completed measures of attachment, alliance and service engagement. Staff completed measures of alliance and service engagement. Psychotic symptoms and perceived racial/ethnic discrimination were measured as potential confounders. Results indicated that higher attachment anxiety was associated with poorer client-rated alliance but not staff-rated alliance, except on the goal agreement component. Attachment avoidance was not related to alliance. Attachment was not associated with service engagement. Attachment ratings were also compared to secondary data collected from a large White sample. Clinical implications are discussed and areas for future research are highlighted. Paper 3 provides a critical appraisal of the design, implementation and interpretation of findings for Papers 1 and 2. Strengths and limitations of both papers are discussed, and personal reflections about the research process are included.