IntroductionLow back pain (LBP) is a ubiquitous health problem. It represents the most frequent illness in humans. LBP lifetime prevalence ranges 60-90%, and the annual incidence is 5% [6]. It is reported to be the leading cause of disability in people under 45 years of age and the third cause of disability in those over 45 [21]. High costs associated with LBP and its socio-economic impact have made this so-called self-limited and benign condition a considerable health-care policy challenge, especially in industrial countries [1,7,20,21]. LBP may have an impact on the functional status of the patient, interfering with basic activities like standing, walking, dressing and many work-related activities, leading to pain and disability. Measuring the disability is important. First, as an outcome measure, it is an important indicator of the of the patient's quality of life. In addition, this outcome can be used for planning and monitoring therapy, clinical and epidemiological reAbstract Our aim was to validate a culturally adapted, Tunisian-language version of the Roland-Morris Disability Questionnaire (RMDQ), which is a reliable evaluation instrument for low-back-pain disability. A total of 62 patients with low back pain were assessed by the questionnaire. Reliability for the 1-week test/re-test was assessed by a construction of a Bland Altman plot. Internal construct validity was assessed by Cronbach's αtest. External construct validity was assessed by association with pain, the Schober test and the General Function Score. Sensitivity to change was determined using a t-test for paired data to compare RMDQ scores at inclusion and at completion of the therapeutic sequence of local corticosteroid injections. We also compared the questionnaire score with the General Function Score, both taken after completion of the therapeutic sequence. The constructed Bland Altman plot showed good reliability. Internal consistency of the RMDQ was found to be very good and the Cronbach's α test was 0.94, indicating a good internal construct validity. The questionnaire is correlated with the pain visual analogue scale (r=33; p=0.0001), with the Schober test (r=0.27; p=0.0001) and the General Function Score (r=56; p=0.0001) indicating an adequate external construct validity. The RMDQ administered after the therapeutic sequence is sensitive to change (r=0.83; p=0.000). Comparison of the questionnaire score to the General Function Score, after completion of the therapeutic sequence, was satisfactory (r=0.75; p=0.000). We conclude that the Tunisian version of the Roland-Morris questionnaire has good reliability and internal consistency. Furthermore, it has a good internal-and external construct validity and high sensitivity to change. It is an adequate and useful tool for assessing low-back-pain disability.