Background: Flavor industry workers exposed to diacetyl in butter flavorings are at risk for bronchiolitis obliterans, manifested mainly by accelerated declines in FEV1 on spirometry associated with worsening respiratory symptoms. We analyzed longitudinal spirometry and symptoms in flavor workers to assess possible work-related fixed airways disease. Methods: 191 workers participated in medical surveillance using serial questionnaires and spirometry. Task-specific diacetyl exposures were measured using NIOSH methods 2557. We used the general linear mixed model to evaluate longitudinal change in FEV1 & respiratory symptoms, then stratified overall changes by smoking status and analyzed by four different exposure variables: job category; reported use of diacetyl,; work with powders, liquids or spray-dried flavors; and cumulative diacetyl exposure. Results: Decline in FEV1 (adjusted for height, race and gender) was not significantly different between production and non-production workers (-40.40 v. -40.67 ml/yr, p=0.95), even when stratified for smoking. In the overall group of flavor workers, we found no significant differences in FEV1 based on reported frequency of diacetyl use, work with different flavor formulations, or cumulative diacetyl exposure. However, never and former smoker production workers in the high cumulative exposure group had greater declines in FEV1 (-41.77 ml/yr and -44.00 ml/yr) compared to current smoker production workers (-29.44 ml/yr, p=0.02 for former v. current smokers). Among participants working in flavor manufacturing for two years or less, workers in the high cumulative diacetyl exposure group did have a greater rate of decline in FEV1 (-65.14 ml/yr) compared to medium ( 42.75 ml/yr, p=0.0166) and low exposure groups (-46.69 ml/yr, p=0.0151). There were few differences in respiratory symptoms or asthma diagnosis among any of the exposure groups. Only for the symptom chest tightness was the yearly prevalence increasing for production workers and workers reporting use of diacetyl, compared to non-production workers (p=0.0069) and those not reporting use of diacetyl (p<0.0001). Conclusions: We found no significant differences in rate of FEV1decline and rate of change in respiratory symptoms over time among flavor workers characterized by four different workplace exposure variables. However, more recently hired flavor workers in the high exposure group did have significantly greater declines in FEV1 compared to the low exposure group, suggesting that regular medical surveillance for work-related lung disease is important, particularly in this subset of flavor workers. We also found that currently smoking flavor workers in high diacetyl exposure categories may have a slower rate of lung function decline compared to non-smokers in the high exposure group, suggesting a “protective” effect from smoking relative to flavor chemical effects.