The Obstacles to Return-to-Work Questionnaire (ORQ) was developed and evaluated. A total of 154 patients with chronic musculoskeletal pain and prolonged work disability participated in the study. Factor analyses reduced the ORQ to 55 items grouped into 9 subscales. The subscales were named "Depression," "Pain intensity," "Difficulties at work return," "Physical workload and harmfulness," "Social support at work," "Worry due to sick leave," "Work satisfaction," "Family situation and support," and "Perceived prognosis of work return." The subscales showed satisfactory reliability. In order to determine predictive validity a discriminant analysis was conducted with sick leave 9 months after assessment as the outcome. This analysis indicated that the scales "Perceived prognosis of work return," "Social support at work," "Physical workload and harmfulness," "Depression," and "Pain intensity" could significantly predict sick leave and correctly classified 79% of the patients. The Multidimensional Pain Inventory and the Disability Rating Index could also significantly predict sick leave in this sample and correctly classified as many patients as the ORQ. However, these questionnaires do not include any work-oriented items and they had a lower specificity than the ORQ. This study suggests that patients' perceptions and beliefs about work and returning to work may be a significant hindrance for actual recovery.