INTRODUCTION: During the last 20 years a great number of studies have emphasized the potential role of psychological factors as relevant predictors for the first onset of back pain as well as for the development of chronic pain. The formulation of a biopsychosocial perspective of the etiology and chronicity of back pain was a natural consequence. Actual questions concern the relative impact of psychological risk factors in the process of chronicity of back pain compared to biomedical, social and occupational factors. METHODS: Whereas several review articles regarding the role of psychological risk factors are available up to now, a recently published review conducted by Steven Linton was the first systematic analysis of well controlled prospective studies published since 1967. Using a grading system similar to that recommended for guidelines the author defined a level A evidence when at least two good-quality prospective studies supported the prospective power of a variable. Level B evidence had support from one prospective study. Level C represented inconclusive data and level D indicated that no studies met the criteria. Based on a literature search of more than 900 studies, 37 good-quality prospective studies were analysed in detail. RESULTS: The results indicated level A evidence for the following interrelations: (1) Psychosocial variables, especially chronic distress in daily life, depression and work dissatisfaction were clearly associated with the onset of back and neck pain (2). Psychological variables, especially chronic stress in daily life, work dissatisfaction, depression and pain-related cognitions and coping behaviour were clearly linked to the transition from acute to chronic pain and disability (3). Psychological variables generally had more prospective power than biomedical, social or objective occupational variables. Among the pain-related cognitions, catastrophizing and fear-avoidance-beliefs yielded the most empirical support. Among coping behaviour passive coping stretegies such as avoiding behaviour was most important. At least level B evidence has been shown for the counterpart of an extreme suppressive coping behaviour. Patients who tended to suppress or ignore pain in order to finish all activities they started, who were unable to integrate phases of passive relaxation into the daily routine displayed a high risk of chronicity of pain 6 months after an acute phase of pain. One international and two German good-quality prospective studies have shown the predictive power of a maladaptive suppressive behaviour pattern. Based on the avoidance-endurance-model of pain chronification, both extreme and one-sided passive pain coping as well as one-sided suppressive coping modes were conceptualized as maladaptive due to the process of chronicity. We assume that extreme passive behaviour will lead to immobility and muscular atrophy. Neurophysiological processes of sensitization will further lead to the development of chronic pain. On the other hand, extreme suppressive behaviour will lead to an overuse of muscles and joints with a repetitive combination of muscular hyperactivity and pain. These repetitive pain experiences will also elicit neurophysiological processes of sensitization. (4) Psychosocial variables displayed more predictive power than biomedical or biomechanical factors. (5) Psychosocial factors may be used as predictors of the risk for developing long-term pain and disability. Mixed empirical evidence has been shown for the role of personality factors (level C evidence) and no support has been found for the idea of a "pain prone" personality disorder or for the role of sexual and physical abuse. DISCUSSION: 20 years of research, several qualitative reviews and a recently published systematic review of 37 good-quality prospective studies regarding the role of psychological, biomedical, social and objective occupational factors in the process of chronicity of back pain revealed that psychological factors are significantly related to the onset of back pain as well as to the development of chronic pain. Furthermore, the psychological factors displayed more predictive power than biomedical or biomechanical variables. As a consequence for clinical practice, these psychological risk factors have to be considered as "yellow flags" if a back pain problem won't respond to medical treatment for more than four weeks. As a further consequence, special risk factor - based psychosocial interventions should be offered in addition to the medical treatment to patients with high psychological risk factors for the development of chronic pain. The efficacy of such an interdisciplinary care in the very early phases of back pain in order to prevent chronicity of pain and disability has been published recently by our group.