Purpose: We assess the effects of racial or ethnic concordance between caregivers and interventionists on caregiver attrition, change in depression, and change in burden in a multisite randomized clinical trial. Design and Methods: Family caregivers of patients with Alzheimer's disease were randomized to intervention or control groups at six sites from 1996 to 2000. Interventionists provided psychosocial interventions aimed at decreasing caregiver depression and burden. This analysis included 694 caregivers who received face-to-face interventions from 36 interventionists at five sites. We modeled caregiver loss to follow-up at 12 months and changes in depression and burden from baseline by racial or ethnic concordance by using random effects logistic and linear regression models, controlling for caregiver age, gender, race or ethnicity, relation to care recipient, interventionist race or ethnicity, and care recipient activities of daily living. The loss to follow-up model also controlled for care recipient institutionalization and death. Results: Overall, there was no difference in caregiver loss to follow-up or change in depression or burden by racial or ethnic concordance. African-American caregivers with an interventionist of the same race or ethnicity had greater decreases in depression than did African-American caregivers with interventionists of a different race or ethnicity. However, this finding has to be interpreted cautiously because there were only two African-American interventionists. Implications: Although these initial findings do not provide conclusive evidence on whether racial or ethnic concordance is associated with intervention outcomes for caregivers of Alzheimer's disease patients, these results, along with the paucity of research studies evaluating this issue, suggest that the effects of racial or ethnic concordance in research should be systematically examined in future studies. (Contains 5 tables.)