The widely disseminated surveillance case definition of the eosinophilia-myalgia syndrome (EMS) recommended by the Centers for Disease Control in 1989 has never been validated by an appropriate challenge and has commonly been used for unintended purposes. To accurately classify patients for clinical and epidemiologic studies and to properly diagnose individual patients, an acceptable set of criteria must be established. In determining which combination of clinical and laboratory manifestations should be properly included in criteria for EMS the following elements should be considered: (1) the presence of a well delineated, documented acute episode with characteristic symptoms, signs, and laboratory abnormalities; (2) the presence of characteristic histopathological abnormalities; (3) the presence of objective evidence for involvement of the most commonly affected major organs: skin, muscle, nerve, and/or lung; and (4) the absence of premorbidities or comorbidities that could explain the components of illness on which the diagnosis is based. A preliminary set of criteria is proposed in this paper. This model requires further scrutiny, revision, and testing.