For decades data have been collected comparing health care in racial a
nd ethnic groups. The use of such groups in health services research a
ssumes that standard, reliable, and valid definitions of race and ethn
icity exist and that these definitions are used consistently. In fact,
race is a term often used, but ill defined. It can incorporate biolog
ical, social, and cultural characteristics of patients and can refer t
o both genetic and behavioral traits. Various investigators have repor
ted differences between racial and ethnic groups in health status, dis
ease manifestation and outcome, resource utilization, and health care
access, often specifying neither a definition of race nor the measurem
ent they used to classify their study populations. The role of race as
an explanatory variable in health services research requires greater
scrutiny than many researchers currently provide. Many studies use rac
e as a proxy for other socioeconomic factors not collected in the rese
arch effort. This article explores the ambiguities about race as an ex
planatory variable that render such research difficult to interpret. W
e suggest that health services researchers focus on nonracial socioeco
nomic characteristics that might be both more informative and more use
ful in guiding policy formation.