The aim of this study is to develop a prospective health-based risk adjustment model for sickness insurance funds in Switzerland which, in addition to age and gender, uses medical information from prior use of health care services. In fact, because information on ambulatory diagnoses is currently not available in Switzerland, we will limit our presentation to risk adjustment models which use diagnostic information from prior hospitalisation. [Authors, p. 9]. [Contents] 1. Introduction. 2. Risk selection and risk adjustment issues. 3. Risk adjustment: general considerations. 4. Econometric modelling. 5. The patient classification systems used in this study. 5.2 Patient classification systems based on AP-DRGs and SQLape. 6. Data and Method. 7. Development of a prospective health-based risk adjustment model: a preliminary. 7.1 Intermediary results for the one-year models. 7.1.1 Results for canton Vaud. 7.1.2 Results for the canton of Zurich. 7.1.3 OLS versus non-linear models. 8. Prospective health-based risk adjustment model. 9. Implementation of the proposed health-based risk adjustment model. 10. Future research.