OBJECTIVE: A legacy of racial discrimination in medical research and the health care system has been linked to a low level of trust in medical research and medical care among African Americans. While racial differences in trust in physicians have been demonstrated, little is known about racial variation in trust of health insurance plans and hospitals. For the present study, the authors analyzed responses to a cross-sectional telephone survey to assess the independent relationship of self-reported race (non-Hispanic black or non-Hispanic white) with trust in physicians, hospitals, and health insurance plans. METHODS: Respondents ages 18-75 years were asked to rate their level of trust in physicians, health insurance plans, and hospitals. Items from the Medical Mistrust Index were used to assess fear and suspicion of hospitals. RESULTS: Responses were analyzed for 49 (42%) non-Hispanic black and 69 (58%) non-Hispanic white respondents (N=118; 94% of total survey population). A majority of respondents trusted physicians (71%) and hospitals (70%), but fewer trusted their health insurance plans (28%). After adjustment for potential confounders, non-Hispanic black respondents were less likely to trust their physicians than non-Hispanic white respondents (adjusted absolute difference 37%; p=0.01) and more likely to trust their health insurance plans (adjusted absolute difference 28%; p=0.04). The difference in trust of hospitals (adjusted absolute difference 13%) was not statistically significant. Non-Hispanic black respondents were more likely than non-Hispanic white respondents to be concerned about personal privacy and the potential for harmful experimentation in hospitals. CONCLUSIONS: Patterns of trust in components of our health care system differ by race. Differences in trust may reflect divergent cultural experiences of blacks and whites as well as differences in expectations for care. Improved understanding of these factors is needed if efforts to enhance patient access to and satisfaction with care are to be effective.