The effects of ethnically matching tutors and mothers participating in Project PRIMER (PRoducing Infant/Mother Ethnic Readers), a community-based literacy program, were investigated. It was hypothesized that mothers ethnically matched to tutors would show increased satisfaction with the program, greater increases in literacy-promoting behaviors, and decreased attrition. It was also hypothesized that their children would show greater progress in conceptual knowledge and language production. Tutors were trained to teach Head Start parents techniques to increase language development in their children. During each instructional visit, tutors introduced literacy tips to mothers, modeled reading, taught an abstract concept, and sang a song with the mother and child. Tutors also took the mother and child on four library visits. Mothers' satisfaction with the tutor was assessed, and their progress was measured by scales assessing reading, teaching, and library behaviors at both pre-and post-assessments. Children's gains were assessed via standardized tests of conceptual knowledge and language production. The results indicated that there was no advantage of matching tutors and tutees on ethnic similarity. The implications of these findings for the use of indigenous paraprofessionals is discussed.