Existing self-report measures for measuring the severity of a depression are not very specific for the disorder concerned. Especially the low divergent validity with regard to anxiety disorders is a problem; a number of reasons are clarified in this article. Attention is paid to a theoretical model concerning the role of positive and negative affectivity in depression and anxiety. This model offers an additional opportunity for differentiating both disorders based on specific profiles of affectivity. A self-report measure for measuring affectivity (the PANAS) is presented. Concluding, a report will follow of research with a Dutch translation of this measure with healthy volunteers and depressed patients. The reliability and validity appeared satisfactory. In addition a depression specific profile could be replicated in a group of depressed patients.