BACKGROUND: Previous research suggests that lack of knowledge of hospice is a barrier to the use of hospice care by African Americans. However, there is little data examining racial differences in exposure to hospice information. OBJECTIVES: Examine racial differences in self-reported exposure to hospice information and determine how this exposure impacts beliefs about hospice care. METHODS: We surveyed 200 community-dwelling older adults (65 or older). We used Spearman's correlations to examine the relationship between responses to individual items on the Hospice Beliefs and Attitudes Scale (HBAS) and self-reported exposure to hospice information (never heard of hospice, heard a little, or heard a lot). We used multivariate analyses to examine predictors of exposure to hospice information and beliefs about hospice care (total score on HBAS). RESULTS: Compared to whites (n = 95), African Americans (n = 105) reported significantly less exposure to hospice information (p = 0.0004). Nineteen percent of African Americans and 4% of whites had never heard of hospice; 47.6% of African Americans and 71.6% of whites had heard a lot about hospice. In multivariate analysis controlling for demographics and health status, African Americans had a two times higher odds of reporting that they had never heard of hospice or heard only a little about hospice versus heard a lot about hospice (odds ratio [OR] = 2.24 [1.17, 4.27]. Greater exposure to hospice information was associated with more favorable beliefs about hospice care (outcome: total score on HBAS; parameter estimate 1.34, standard error 0.44, p = 002). CONCLUSIONS: African Americans reported less exposure to information about hospice than whites. Greater exposure to hospice information was associated with more favorable beliefs about some aspects of hospice care. Because knowledge is power, educational programs targeting older African Americans are needed to dispel myths about hospice and to provide minorities with the tools to make informed choices about end-of-life care.