The history, reality, and prolonged effects of racial stratification and its supportive ideology in the United States require systematic study of the impact of racial perceptions in the health sphere and in every facet of American life (Berry 1965; Berreman 1985; Blackwell 1985). In this regard, the use of race as an independent variable in social research is, in principle, similar to the presumed explanatory power of other status characteristics such as class or economic position. Depending on which meaning is intended, the potential policy, health services, and sociopolitical outcomes could be diametrically opposed to one another. Far from being an esoteric subject among intellectuals, any definition of race has fundamental and practical extensions to cultural and political realities. Essentially, studies and discussions of racial similarities and differences in health matters, whether intended or not, go beyond statistical compilations and correlations and reflect norms, values, the country's common beliefs (Praeger 1982), and the structural positions of majority and minority groups. Presumably, these are among the reasons that racial categories are studied in social science. They are assumed to represent socially relevant and unique histories, experiences, and statuses which differentiate black and white Americans in particular. The risks in epidemiologic and in social science research involve the preoccupation with disparities in the health difficulties among them; the attribution of racial biology and genetic traits to virtually all health spheres; the assumed preponderance of disabling conditions for blacks; and the unrelenting focus on only two racial populations despite our having a multiethnic society (Wilkinson 1987). Since health behaviors are directly associated with a group's "way of life," they should be carefully scrutinized within relevant socio-environmental contexts as part of the scientific processes of discovery, for the underlying premises, ideological translations, and practical applications of their studies especially with respect to race-specific health research. It is likely that systematic probing beyond demographic or constitutional factors will enable social scientists and health researchers to discover that for certain behaviors (e.g., prevention), individual attributes such as race and sex--and even knowledge, roles, attitudes, and diets--may explain far less than will environmental hazards and basic structural variables such as the organization of the health care delivery system, availability of and access to care, ability to pay, provider patterns, diagnostic processes, institutional operations, and quality of care.(ABSTRACT TRUNCATED AT 400 WORDS)