A discussion was presented of disease characteristics and exposure considerations with respect to epidemiologic studies of disease etiologies. Definitions of disease entities and diagnoses were discussed. It was noted that inclusion of an exposure in diagnostic criteria would bias further research on associations of exposures with the disease. Precise disease characterization was often important for obtaining precise risk information. In some cases, definition by histology, anatomic site or particular subtype of the disorder provided greater specificity. Conversely, merging of various disorders could also be helpful, particularly for ill defined disorders and complex exposure situations. Some relationships of high disease risk with certain occupational exposures in individuals with metabolic or genetic predispositions were identified, including genetic alterations associated with carcinogenesis. Exposure assessment could be problematic in terms of actual exposure and relevant exposure characteristic. Time elements were discussed, including latency, early versus later exposures, acute versus long term exposures, relative importance of earlier and later exposures when both occurred and parental or early life exposures. Aging and environmental exposures were possible contributors to latency effects. Because exposure assessment variables were often numerous, analysis could be complex. Development of exposure blocks (combinations of variables) was discussed as a way of approaching such complexity. An example of this was use of job title as an exposure surrogate. Exposure blocking could assist in controlling confounding factors and in studying interactions. Ethical considerations of identification of susceptible phenotypes were noted. The authors suggest that such information be used for improvement in work environment safety rather than discrimination in hiring.