I n 2003, the Institute of Medicine issued a groundbreakingreport that shone a spotlight on the problem of racial and ethnic disparities in health care in the United States. 1 It is now well recognized that minority Americans tend to receive less care, and lower-quality care, than the majority (white) population. The reasons behind this phenomenon are complex and opaque. There is no single explanation, and in many cases, we do not have an explanation. However, for millions of minority Americans, most of them foreign-born, one explanation is readily apparent: they do not speak, read, or write English, and health care in this country, with few exceptions, is delivered in English.Two articles in this JGIM Supplement attest to the role of language barriers in explaining racial/ethnic disparities in health care. Cheng and colleagues examined the use of basic, evidence-based health care services for prevention and chronic disease management among Latinos and whites in the United States. 2 They found that whites were more likely than Latinos as a whole to receive all recommended services for which they were eligible. However, when grouped by language use, Latinos who spoke English at home were not significantly less likely than whites to receive recommended services, whereas those who spoke Spanish at home were, even after accounting for education, insurance coverage, usual source of care, and a host of other potential explanatory factors. Thus, the bulk of the racial/ethnic disparity between whites and Latinos, in this case, is actually a "language" disparity, between English-speakers and Spanish-speakers.The study by Sentell et al. produced similar findings. 3 They examined access to mental health care in California, the state with the highest proportion of limited English-proficient (LEP) persons in the country. 4 They found that among individuals stating a need for mental health services, Latinos were less likely than whites to have received them. However, Englishspeaking Latinos were similar to whites in their access to mental health care, whereas non-English-speaking Latinos had a markedly lower likelihood of receiving services (even when they had mental health insurance coverage). Again, the disparity here is not so much a racial/ethnic one per se, as one related to English proficiency.The findings of these studies indicate not only that language plays a large role in explaining racial/ethnic disparities, but that comparing groups by race/ethnicity alone may substantially underestimate the degree of inequity between the haves and the have-nots. For instance, in the study by Sentell et al., Latinos as a whole were about one-third less likely to receive needed mental health services when compared to the overall California population (28% vs 45%). But Latinos who spoke no English were 85% less likely to receive services compared to the general, English-speaking population (8% vs 51%). 3 In the same vein, examining groups by English proficiency can reveal "hidden" disparities. Sentell and colleagues found that Asian...