OBJECTIVE Four diacetyl workers were found to have bronchiolitis obliterans syndrome. Exposures, respiratory symptoms, lung function and exposure-response relationships were investigated. METHODS 175 workers from a plant producing diacetyl between 1960 and 2003 were investigated. Exposure data were used to model diacetyl exposure. Lung function and questionnaire data on respiratory symptoms were compared to a general population sample and respiratory symptoms to an internal reference group. RESULTS Workers were potentially exposed to acetoin, diacetyl, acetaldehyde and acetic acid. Historic diacetyl exposure ranged from 1.8 to 351 mg/m(3), and from 3 to 396 mg/m(3) for specific tasks. Diacetyl workers reported significantly more respiratory symptoms compared to the general population sample (continuous trouble with breathing (prevalence ratio (PR) = 2.6; 95% CI 1.3 to 5.1), daily cough (PR = 1.5; 95% CI 1.1 to 2.1), asthma attack (ever) (PR = 2.0; 95% CI 1.2 to 3.4), doctor diagnosed asthma (PR = 2.2; 95% CI 1.3 to 3.8) and asthma attack in the last year (PR = 4.7; 95% CI 1.9 to 11.4)) and to a minimally exposed internal reference group (ever trouble with breathing (PR = 2.8; 95% CI 1.1 to 7.0) and work-related shortness of breath in the last year (PR = 7.5; 95% CI 1.1 to 52.9)). Lung function did not differ between groups. A positive relationship between exposure and FEV(1) was found. CONCLUSIONS The excess of respiratory symptoms in this retrospective cohort suggests that diacetyl production poses an occupational hazard. Limited historical exposure data did not support a quantitative individual diacetyl exposure-response relationship, but our findings suggest that preventive measures are prudent.