Objectives: This study explored how patients with back and knee pain perceived a home exercise programme consisting of isometric and isotonic components using three different prescription methods - brochure, verbal and non-verbal videotapes. Patients and methods: Thirty patients with non-specific low-back pain and knee pain were referred to the physical therapy department for exercise therapy. Subjects were randomly assigned to brochure group, non-verbal group, and verbal videotape group. Each group performed the isometric and isotonic exercises according to assigned instructions. An Exercise Assessment Scale was also developed to measure exercise performance for this study. Results: All demographic parameters were homogeneous in three groups. Isotonic exercises were more effective than isometric exercises in all three prescription methods. The Exercise Assessment Scale showed that the starting positions to both isometric and isotonic exercises were the same and correct in all three prescription conditions. Range of motion and planes of movement were perceived better by the patients during isotonic exercises than those of isometric exercises. In all groups, isotonic exercise compensation scores were higher than the isometric exercises scores. Conclusions: Verbal videotape exercise instructions and dynamic modelling (videotaped) were slightly better than other instructions and static modelling (brochure). The instruction of therapeutic exercises with periodic supervision in physiotherapy out-patient clinics appears to be adequate in assisted home exercise programmes. © W.S. Mimey & Son Ltd 2007.