In March 2008, an International Brotherhood of Teamsters local union requested a health hazard evaluation at a flavoring manufacturing company because of concern about possible lung effects of flavoring exposures, including diacetyl. In May 2008, NIOSH conducted a brief walk-through visit of the company. Subsequently, the company brought litigation to prevent further on-site evaluations. In June 2009, after resolution of the litigation, the company agreed to provide NIOSH with various medical and workplace information. All information was received by November 2009. NIOSH staff evaluated spirometry data supplied by the company on production employees tested from 1998 through 2009. The majority of the spirometry tests was performed in the years 2004 and later. We classified spirometry tests for quality and compared the prevalence of abnormalities in acceptable quality tests to national population prevalence, adjusted for the distributions of age, sex, race, smoking, and body mass index in the company's employees. We calculated declines over time in forced expiratory volume in 1 second (FEV1), a lung function measurement made using spirometry, for employees with more than one spirometry test of acceptable quality, since excessive decline in FEV1 can be an early marker of lung disease. Using an approach that adjusts for quality of a spirometry monitoring program, reflected by within-person variation, we compared the declines in lung function to a statistically-determined lower limit of normal decline. After establishing which employees had abnormal declines in lung function, we evaluated work area risk factors for associations with excessive declines in lung function, adjusted for age, smoking, tenure, change in weight, and obesity. Based on our experience in other flavoring plants, we designated a group of areas with higher potential for flavorings exposure as a possible risk factor. These were liquid compounding, process flavors, dry blend, extract and distillation, and spray dry. We compared spirometric findings for employees in these areas with employees who worked in other areas of the plant. We also evaluated environmental monitoring measurements supplied by the company before and after our walk through visit, along with the four measurements conducted during the walk through. The flavorings manufacturing company supplied spirometry data on 112 employees; 75 percent of these employees had more than one test session, with a follow-up range of 0.5 to 11 years. The most recent spirometry measurement for 106 employees with at least one spirometry test of acceptable quality showed that 30/106 (28 percent) employees had restrictive abnormalities, 3/106 had obstructive abnormalities, and 1/106 had both restrictive and obstructive abnormalities. The 28 percent of employees with restrictive abnormalities was 3.8 (95 percent confidence interval 2.6-5.4) times higher than would be expected in a U.S. population with the same demographic characteristics. Among the 30 employees with restrictive abnormalities, 27 percent had longitudinal testing demonstrating abnormal declines in FEV1 over time, indicating that progressive deterioration in lung function had occurred; 17 percent had no history of longitudinal testing, and thus no evaluation for excessive decline over time was possible. In addition to the 34/106 (32 percent) with abnormal restrictive or obstructive spirometry, 5 employees with normal most-recent spirometry values had longitudinal testing demonstrating abnormal declines in FEV1 over time, which, if continued, might result in spirometric abnormality. Thus a total of 39/106 (37 percent) employees among those with company spirometry measurements had evidence of some abnormality, either in classification of most recent spirometry as showing restriction or obstruction; and/or longitudinal spirometry showing excessive decline over time, with most recent spirometry values still within the normal range. Forty-two employees did not have serial data of adequate quality to allow evaluation for abnormal declines. The company's 2009 exposure measurements documented that diacetyl was present in at least one sample in all sampled production areas, the laboratory, and the warehouse. Two of the four NIOSH area measurements detected diacetyl in liquid compounding and packaging areas. These findings supported our designation of a group of areas with higher potential for flavorings exposures, although some areas that were classified as being in the lower potential for exposure group did have exposures of concern. Employees who ever worked in areas with higher potential for flavorings exposure (liquid compounding, process flavors, dry blend, extract and distillation, and spray dry) had 2.8 times greater average annual declines in FEV1 than employees in areas with lower potential for flavorings exposure. In particular, employees who had ever worked in liquid compounding had statistically higher average annual declines in FEV1, compared to the lower potential for exposure group. Employees who currently worked in higher potential for flavorings exposure areas were 7 times more likely to have abnormal declines in FEV1 than employees in other areas, which is consistent with work-related risk for adverse respiratory health outcomes. Additional medical testing is needed to determine whether the abnormal spirometry findings found among workers are due to lung disease. In particular, medical testing is needed to determine the underlying cause of restrictive spirometry. Although obesity is a major cause of restrictive abnormalities in the United States, our comparisons with the U.S. population were adjusted for the proportions of employees who were overweight or obese, as were our analyses of work area risk factors. Thus, restrictive spirometry in this workforce cannot be explained by obesity. Since the excess of spirometric abnormalities is substantial and the distribution of excessive declines in lung function is associated with history of working in areas with higher potential for exposure, there is great cause for concern about occupational lung disease. It is possible that some exposure other than diacetyl may underlie these abnormalities, since the predominant pattern of restrictive abnormalities differs from the pattern of obstructive abnormalities seen among microwave popcorn employees. Also, the flavorings used in this plant are more diverse than are found in microwave popcorn production. However, some diacetyl-exposed individuals in microwave popcorn plants and other settings have had restrictive abnormalities without other apparent cause. Thus, the spectrum of abnormalities caused by diacetyl might include restrictive lung disease. This possibility remains to be fully explored. We recommend further lowering of flavoring exposures, without regard to anticipated exposure limits to diacetyl, since other chemicals may be associated with the adverse respiratory health outcomes documented in the workforce. We reiterate our interim report recommendations for engineering controls, work practices, enhanced respiratory protection, and medical surveillance. Ongoing medical surveillance that uses longitudinal spirometry testing to monitor those with potentially harmful exposures is also recommended, particularly until the cause of the high burden of abnormal spirometry in the workforce is fully understood. The company's contract with its medical provider should provide for aggregate epidemiologic analysis of the medical results, including analysis of medical results by department or job. Aggregate analysis can identify hazards associated with flavoring manufacture and may assist the company in targeting priorities for prevention through exposure control.