This thesis focuses on two specific clinical areas: music therapy with pre-school children with autistic spectrum disorder and their parents, and music therapy diagnostic assessments with children between the ages of four and twelve who are suspected of being on the autistic spectrum. Firstly, the literature was examined and the clinical work was described in detail. This process made it possible to determine what characterises the author’s particular approach, and to find out how it may be different to other music therapists’ work. A 45 minute video which illustrates the approach with pre-school children with autism and their parents accompanies this thesis. Two outcome research investigations were carried out. The first involved studying ten pre-school children with autistic spectrum disorder and their parents who received weekly, individual music therapy sessions over a period of 18 to 26 weeks each. The sessions were video-taped and the videos analysed in detail. The parents were interviewed and asked to fill in questionnaires both pre- and post-treatment. Nine out of the ten dyads achieved some or all of the individual aims set out before treatment began. The parents all felt that music therapy had been effective. The author also looked at how she spent her time in music therapy sessions across the ten children and found that she was generally very active and spent a high proportion of her time vocalising. The second investigation compared Music Therapy Diagnostic Assessments (MTDA) with Autistic Diagnostic Observation Schedules (ADOS) carried out on 30 children suspected of being on the autistic spectrum. A scoring system similar to that used for the ADOS was devised for the MTDA especially for this research investigation. In addition, the children were interviewed after both the MTDA and the ADOS and the people carrying out the tests filled in a questionnaire about their perceptions of the assessment tool after every test. The two assessments showed 72 % of agreement between diagnostic categories, indicating that the MTDA was providing similar information as a recognised and established diagnostic tool. However, the two assessments also showed significant differences in scores of individual questions, indicating that the MTDA could serve a useful and distinct purpose in helping the psychiatric team to diagnose children with autism. The children generally enjoyed the assessments and the music therapist felt that the test was easy to carry out and score, indicating that the MTDA was ‘user-friendly’. Throughout this thesis the author has adopted a personal style particularly when describing her own clinical work and when examining the literature. Although the two outcome investigations rigorously examined numerical data, the author also described her own impressions as the research investigation progressed.