Objective: Multimodal rehabilitation programmes (MMRP) for chronic pain could be improved by determining which patients do not benefit fully. General distress and pain-related fear may explain variations in the treatment effects of MMRP. Design: Cohort study with a cross-sectional, prospective part. Patients: Chronic musculoskeletal pain patients referred to 2 hospital-based pain rehabilitation clinics. Methods: The cross-sectional part of this study cluster analyses patients (n = 1,218) with regard to distress and pain-related fear at first consultation in clinical pain rehabilitation and describes differences in external variables between clusters. The prospective part follows the subsample of patients (n = 260) participating in MMRP and describes outcome post-treatment. Results: Four distinct subgroups were found: (i) those with low levels of distress and pain-related fear; (ii) those with high levels of pain-related fear; (iii) those with high levels of distress; and (iv) those with high levels of distress and pain-related fear. These subgroups showed differences in demographics, pain characteristics, quality of life, and acceptance, as well as the degree of MMRP participation and MMRP outcome. Conclusion: Among patients with chronic pain referred to MMRP there are subgroups with different profiles of distress and pain-related fear, which are relevant to understanding the adaptation to pain and MMRP outcome. This knowledge may help us to select patients and tailor treatment for better results.