The multidimensional and heterogeneous etiogenesis of chronic low back pain requires multimodal as well as specific treatment programmes. Aim of the present study was to increase the specificity of treatment already at the interface "family health care - assigning institution - rehabilitation clinic" by supporting the assignment to orthopaedic or psychosomatic rehabilitation through extensive diagnostics and deriving a patient's profile of risk factors in the forefield of rehabilitation. In a prospective randomized control group design the rehabilitation success in 309 persons (age 22 - 58) suffering from chronic low back pain was compared on risk factor specific vs. conventional assignment. Between 22 % and 27 % of the assignments following conventional criteria have to be considered as false. On the other hand, rehabilitation success did not turn out to have been increased by specific assignment. In conclusion, implications for more far-reaching change in treatment programmes within and subsequent to stationary rehabilitation are discussed.