STUDY DESIGN: Cross-sectional investigation and follow-up of patients with low back pain. OBJECTIVES: To evaluate the capacity of the INTERMED--a biopsychosocial assessment and classification system for case complexity--to identify patients with a chronic, disabling course of low back pain and to predict treatment outcome. SUMMARY OF BACKGROUND DATA: An impressive number of biologic and nonbiologic factors influencing the course of low back pain have been identified. However, the lack of a concise, comprehensive, reliable and validated classification system of this heterogeneous patient population hampers preventive and therapeutic progress. METHODS: The INTERMED was used to assess patients with low back pain, who participated in a functional rehabilitation program (n = 50) and patients with low back pain who applied for disability compensation (n = 50). Patients of the rehabilitation program were observed to assess the effects of treatments. RESULTS: The INTERMED distinguished between patients in different phases of disability and provided meaningful information about the biopsychosocial aspects of low back pain. In hierarchical cluster analysis two distinct clusters emerged that differed in the degree of case complexity and treatment outcomes. CONCLUSIONS: This first application of the INTERMED indicates its potential utility as a classification system for patients with low back pain.