OBJECTIVES: To validate the Portuguese version of the Nordic Musculoskeletal Questionnaire (NMQ), and to evaluate the relationship between musculoskeletal morbidity and demographic, occupational and behavior variables. METHODS: Ninety government bank employees were interviewed using the Portuguese version of NMQ in Brasilia, Brazil, in 1999. The respondents' answers were compared to clinical data found in their medical records. Descriptive and associative variable analyses were carried out. Comparative statistical analysis was performed using Student t-test and Pearson's test was applied for correlational analysis. RESULTS: The results revealed an 86% agreement rate between symptoms reported in the NMQ and the respondent's clinical history. There were found differences in symptoms prevalence for the variables gender, occupation and physical activity. Women showed a higher severity average of symptoms for almost all body anatomical regions; managers reported greater severity of symptoms in the lumbar area than clerks; physical activity was associated to the lowest severity of symptoms in upper extremities. CONCLUSIONS: The results show that the Portuguese NMQ version includes a good validity indicator and is satisfactory as a musculoskeletal morbidity measurement tool. However, there is a need of (1) more items to measure the severity of symptoms; and (2) changes in scale layout in order to make it more easily understandable and less prone to avoidable missing data.