This study was undertaken to establish that in an audience situation, music may facilitate an immediate attitude change toward a therapist by patients. To determine the characteristics of such a change, recorded (popular and classical) and live (cello) music was employed. A semantic differential was used to measure attitude change by psychiatric inpatients and control subjects (medical students and student nurses). Significant attitude change were found when the music presented was the preference of the therapist and when this preference was conveyed to the audience. Patient breakdown into diagnostic categories also showed that patients with affective or alcoholic disorders showed significantly larger attitude change than the controls. Results are discussed in terms of cognitive consistency theories of attitude change, concluding that maximal attitude change toward a therapist occurs under conditions in which the therapist presents new aspects of behavior (in terms of previous exposure) to the patients.