Background: In primary care, up to 74% of physical symptoms is classified as unexplained. These symptoms can cause high levels of distress and healthcare utilization. Cognitive behavioral therapy has shown to be effective, but does not seem to be attractive to patients. An exception herein is a therapy based on the consequences model, which distinguishes itself by its labeling of psychosocial distress in terms of consequences rather than as causes of physical symptoms. In secondary care, 81% of the patients accepts this therapy, but in primary care the outcome is poor. We assume that positive outcome can also be reached in primary care, when the consequences model is modified and used bottom-up in an easily accessible group training, in which patients are relieved of being blamed for their symptoms. Our aim is to investigate the (cost-) effectiveness of this training. Methods and design: A randomized controlled trial is designed. One hundred patients are randomized to either the group training or the waiting list. Physicians in general practices and outpatients clinics of general hospitals refer patients. Referral leads to inclusion if patients are between 18 and 65 years old, understand Dutch, have no handicaps impeding participation and the principal DSM-IV-TR classification is undifferentiated somatoform disorder or chronic pain disorder. In contrast to other treatment effect studies, the co-morbidity of a personality disorder does not lead to exclusion. By this, we optimize the comparability between the study population and patients in daily practice enlarging the generalization possibilities. Also in contrast to other effect studies, we chose quality of life (SF-36) instead of physical symptoms as the primary outcome measure. The SF-6D is used to estimate Quality Adjusted Life Years (QALYs). Costs are measured with the Trimbos/iMTA Questionnaire for Costs associated with Psychiatric Illness. Measurements are scheduled at baseline, after the training or waiting list, three and twelve months after the training. The differences between measurements are analyzed according to the intention-to-treat principle. The cost-effectiveness is expressed as costs per QALY, using multiple sensitivity analyses on the basis of a probabilistic model of the trial. Discussion: If we show that our group training is (cost-) effective, more patients could be served, their quality of life could be improved while costs might be reduced. As the training is investigated in a heterogeneous patient group in the daily practice of a mental healthcare institution, its transfer to practice should be relatively easy.