Despite disparities in health care access and quality, African Americans are underrepresented in many areas of clinical investigation, including research in end-of-life care. Because of the importance of surrogate reports in assessing the quality of end-of-life care, this study examined racial differences in next-of-kin participation in an ongoing study of satisfaction with end-of-life care. The parent study includes after-death interviews with next-of-kin of elderly African Americans and Caucasians who died at Duke Hospital. This analysis included next-of-kin of elders who died at Duke Hospital from December 1, 2003 to December 31, 2004. During this period, there were 471 decedents whose next-of-kin were eligible for participation. Of these, 133 (28%) were African American and 338 (72%) were Caucasian. There were no racial differences in completion, contact, or overall response rates. Of those contacted, 39.8% of African Americans and 37.8% of Caucasians completed the study. In multivariate analysis, only the relationship of the next-of-kin to the decedent was an independent predictor of study completion. Children of decedents were significantly more likely to participate than spouses (odds ratio [OR] 2.1 [1.14, 3.86]). In this analysis, next-of-kin of African American and Caucasian decedents were equally likely to participate in an after-death interview assessing satisfaction with end-of-life care. The use of racially concordant interviewers, subject identification with the institution, and the absence of socioeconomic constraints may partly explain these findings. Given the growing diversity of the U.S. population, researchers in end-of-life care must use strategies aimed at recruiting racially and ethnically diverse samples.