A randomized and prospective study examined the effects of intensive communication between treatment providers on physician behavior and patient care using a representative sample of patients referred to a German psychosomatic consultation-liaison (C-L) service (1998-1999). Sixty-seven patients were grouped (ICD-10) and randomized into intervention (n=33) and control groups (n=34). In the control group (CG), only the inpatient referring physicians were informed of consultation findings, as is standard practice in Germany. In the IG, the consultant directly reported the consultation findings to the general practitioner (GP). Patients were examined at 6-month (T2) and 3.5-year (T3) follow-up. Physician feedback and requests regarding the nature of communication were also assessed at T2. A significant reduction in symptoms was revealed at both follow-ups for the IG and CG. No significant group differences were found for acceptance of psychotherapy (59.1% IG and 42.3% CG at T3). The utilization of medical services remained stable across time. Patients who began psychotherapy were shown to have greater baseline symptom levels and "openness for new experiences" (NEO Five-Factor Inventory, NEO-FFi). Regular GP integration of psychosocial aspects into primary care differed between IG (44.0%) and CG (25.0%). Ninety-one percent of the GPs requested the option for a telephone conversation with the consultant or a case conference involving the patient at either their practice (37.7%) or the clinic (24.5%). We conclude that GPs are interested in a differentiated and more intensive cooperation concerning the integration of psychosocial aspects in their outpatient treatment. The patient's acceptance to follow a recommendation for psychotherapy, however, is related to the level of symptom severity and openness for new experiences.