The analysis and validation of prognostic factors is a decisive prerequisite with regard to the development and application of risk-adjusted therapies. Among others, the Sokal index 1 and Kantarjian's new staging system 2 were scores set up to assess the survival of every patient individually. By means of 504 Ph- and/or bcr-abl-positive patients of the German CML study I, we tried to validate both scores mentioned above. Whereas Kantarjian's staging system provided rather unsatisfactory results, the discrimination of different risk groups was more promising with respect to the Sokal index. Still, concerning the differentiation between risk groups for the interferon-alpha (IFN-alpha) treated patients of the German study I, the Sokal index also failed. Hence, we applied Cox stepwise regression technique and identified three prognostic factors for the IFN-alpha treated patients, i.e. eosinophils, erythroblasts (both in peripheral blood), and extramedullary manifestations. We developed our own score and found an encouraging 5-y survival rate of 90% (95% confidence interval (CI) 76-100%) for a certain low-risk group of 41 out of 120 IFN-alpha treated patients. With the need to validate our score and, if possible, not only to confirm the above result but also to identify other particular patient groups, we launched the European Prognostic Factors Project (E.P.F.P.). Within this project, our special interest will be dedicated to patient groups who might also benefit from a bone marrow transplantation.