Aim: The purpose of this study was to investigate the psychometric properties of the EQ-5D-3L and the EQ-5D-5L in patients with back and neck pain. Materials and methods: Participants from two independent clinical survey studies was included in this study. In total 164 participants were answering either the EQ-5D-3L or the EQ-5D-5L after their visit to a physiotherapist in primary care for back and neck pain. Rasch analysis was performed to measure the psychometric properties of the two instruments. Main findings: Overall, the EQ-5D-5L showed preliminary evidence of good psychometric properties. The items in both the EQ-5D-5L and the EQ-5D-3L showed acceptable goodness-of-fit indicating unidimensionality for both instruments measuring the concept health-related quality of life. All test persons fit the model, but the person separation reliability and person separation index were only barely acceptable for the EQ-5D-5L. Conclusions:The results indicate that the EQ-5D-5L should be the preferred option when evaluating health-related quality of life with the EQ-5D, especially for patients with back and neck pain.