Background. The eosinophilia—myalgia syndrome is a newly recognized illness that has been associated with the consumption of tryptophan products. It is not known whether the cause is related to the tryptophan itself or to chemical constituents introduced by the manufacturing process. Methods. To describe the epidemiology of the eosinophilia—myalgia syndrome further and elucidate a possible association with the manufacturing process, we conducted surveillance for the syndrome in Minnesota, a community survey of tryptophan use in Minneapolis-St. Paul, and a case–control study to assess potential risk factors, including the use of tryptophan from different manufacturers. We performed high-performance liquid chromatography on tryptophan samples to identify other chemical constituents. Result. The prevalence of tryptophan use increased from 1980 to 1989 and was highest among women. Among the subjects for whom the source of the tryptophan was known, 29 of 30 case patients (97 percent) and 21 of 35 controls (60 percent) had consumed tryptophan manufactured by a single company (odds ratio, 19.3; 95 percent confidence interval, 2.5 to 844.9; P<0.001). This company used a fermentation process involving Bacillus amyloliquefaciens to manufacture tryptophan. Analysis of the manufacturing conditions according to the retail lot demonstrated an association between lots used by case patients and the use of reduced quantities of powdered carbon in a purification step (odds ratio, 9.0; 95 percent confidence interval, 1.1 to 84.6; P = 0.014), as well as the use of a new strain of B. amyloliquefaciens (Strain V) (odds ratio, 6.0; 95 percent confidence interval, 0.8 to 51.8; P = 0.04). There was a significant correlation (r = 0.78, P<0.001) between the reduced amount of powdered carbon used during manufacturing and the use of the new bacterial strain. High-performance liquid chromatography of this company's tryptophan demonstrated one absorbance peak (peak E) that was present in 9 of the 12 retail lots (75 percent) used by patients and 3 of 11 lots (27 percent) used by controls (odds ratio, 8.0; 95 percent confidence interval, 0.9 to 76.6; P = 0.022). Conclusion. The outbreak of the eosinophilla—myalgia syndrome in 1989 resulted from the ingestion of a chemical constituent that was associated with specific tryptophan-manufacturing conditions at one company. The chemical constituent represented by peak E may contribute to the pathogenesis of the eosinophilia—myalgia syndrome, or it may be a surrogate for another chemical that induces the syndrome. (N Engl J Med 1990; 323: 357–65.).