Usually mental disorders do not occur isolated from each other, but combined. Psychotherapy research in so far as it orientates itself by the paradigm of randomized controlled trials has widely ignored this problem until now. In the present study the attempt is made, to identify combined (complex) mental non-psychotic disorders empirically and to describe them by using collected data from clinical practice. For that purpose, a large unselected sample of inpatients was studied who were treated in the clinic of Tiefenbrunn near Goettingen. On the basis of ICD-10 diagnoses we examined which mental disorders were regularly combined with each other. The ten most common complex mental disorders were identified which cover 75% of the total sample. For these "top ten" we assessed indicators of the severity of disorder chronic suicidally, suicide attempts, inpatient psychiatric pre-treatment, lack of qualifications and admission symptomatology (SCL-90-R, GSI). In addition to characterize these groups we use the pre-post change of the global severity index of the SCL-90-R (GSI) and target problems identified by a goal attainment scaling (GAS). According to the results, there are remarkable differences between these identified complex mental disorders. Complex psychiatric disorders put high demands on both psychotherapy and psychotherapy research. This study contributes to the discussion of these problems. However, further studies are required to corroborate the results.