Smelter and mine workers (1110) responded to an interviewer-administered questionnaire on symptoms and attitudes and performed pulmonary function tests measuring FEV1 and FVC. The correspondence between self-reported respiratory symptoms and measured pulmonary function was found to be significantly related to three psychosocial factors: lack of hypochondriasis, high job satisfaction, and low life stress. All three factors were significantly related to each other. The effects of age, smoking history, work location and years of work (reflecting exposure to sulfur dioxide) were also tested. Ninety-six per cent of the subjects whose symptoms did not correspond with their measured pulmonary function overreported symptoms relative to impairment. Over-reporting was significantly related to low job satisfaction. These findings support the authors' hypothesis that some psychosocial factors may provide useful indicators of the validity of medical questionnaires.