Epidemiologic studies attempt to understand the distribution and determinants of human disease. Epidemiologic research often incorporates information about race, ethnicity, or ancestry, usually as a self-identified race or ethnicity (SIRE) variable. Differences in the distribution and determinants of disease on the basis of SIRE may be identified in these studies. In addition, genetic and other biologic differences according to SIRE are frequently reported. If these differences are real and meaningful, they may have value in identifying disease-causative or -preventive factors, and thus may be beneficial to human health. However, the concepts of race, ethnicity, or ancestry are often poorly considered or crudely applied, particularly in genetic studies of disease etiology or outcome. Consequently, results suggesting genetic differences with respect to disease etiology or outcome across SIRE groups may not be meaningful; in fact, these differences may prove harmful if they propagate stereotypes or spurious differences. Therefore, it is critical to properly consider the meaning, definitions, and use of race, ethnicity, or ancestry in molecular epidemiologic studies.