The long-term effect (1-year-follow-up) of an inpatient client-centred psychotherapy programme was evaluated in 202 severe and chronically disturbed patients with a bread diagnose range (ICD-10 F1-F6). These 202 patients represented 74% of a total of 272 patients who had originally been examined and tested at admission and had now been re-identified after a follow-up of one year. Patients were investigated via Clinical Global Impressions (CGI), Bech-Rafaelsen-Melancholia-Scale (BRMES) and the Personality Inventory (Giessen-Test). Severity of illness, depressivity and self-and-other acceptance improved significantly; social incompetence was especially reduced in patients who had personality disorders. The range of effects reached from good to very good improvement: the magnitude of effect on social variables is probably even under estimated due to high pre-therapy variance. At follow-up 68% of the patients were in occupation or professional education. 32% needed no further treatment. Readmission rate was 14%. Results are discussed with regard to specific client-centred therapy.