We introduce a diagnostic map that was calculated by robust non-metric multidimensional scaling based on AMDP symptom profiles of patients with schizophrenic and affective disorders to demonstrate a possibility to combine the categorical and the dimensional perspective at the same time. In the diagnostic map, a manic, a depressive, and a non-affective cluster clearly emerged. At the same time, the mania dimension (r = 0.82), the depression dimension (r = 0.68), and the apathy dimension (r = 0.74) showed high multiple regression values in the map. We found substantial overlaps of the diagnostic groups with regard to the affective spectrum but irrespective of the ICD-10 classification. Within this sample, we found the association and quality of mood symptoms to be a structuring principle in a diagnostic map. We demonstrate that this approach represents a promising way of combining the categorical and the dimensional perspective. As a practical implementation of these findings, a multidimensional diagnostic map could serve as an automated diagnostic tool based on psychopathological symptom profiles.