Self-esteem has been the focus of a great deal of psychological research due to its many associations with well-being (Bowker, Gadbois, & Cornock, 2003; Wild, Flisher, Bhana, & Lombard, 2004). On average, self-esteem levels decrease markedly during the adolescent years (Birndorf, Ryan, Auinger, & Aten, 2005; Robins & Trzesniewski, 2005). However, adolescent girls tend to experience greater declines in self-esteem than boys, yet the self-esteem of African American girls tends to remain stable (Gray-Little & Hafdahl, 2000; Robins & Trzesniewski, 2005). Developmental (e.g., Erikson, 1968; Shaffer, 2005), gender socialization (e.g., Gilligan, 1982, Turkel, 2000), and cultural diversity research (e.g., Gray-Little & Hafdahl, 2000; Twenge & Crocker, 2002) provide frameworks for understanding the differential changes in self-esteem between adolescent subgroups. However, prior research has not sufficiently considered the potential impact of role models on adolescent self-esteem, or whether the type of role models selected may be impacted by subgroup values. The goals of the present study were to examine whether a relationship existed between type of role model selected and self-esteem levels, and whether characteristics of chosen role models varied, across Caucasian and African American adolescent females. Data for other participants were reported for exploratory purposes. Participants included 53 seventh and eighth grade students. The Rosenberg Self-Esteem Scale (RSE), the Hoped For Possible Selves Questionnaire (PSQ), and a role models form and demographics questionnaire designed by the researcher were utilized. No hypotheses were supported, though results yielded several interesting findings. One-way ANOVAs were performed to compare self-esteem scores and frequencies of role model types. There were no differences in self-esteem across gender or ethnicity, though data trended in the predicted directions. Results additionally showed no differences across ethnicity in frequency of role model selection by type. Pearson r correlations showed that self-esteem scores were positively correlated with selection of distal role models, and negatively correlated with selection of community role models. Pearson r correlations revealed no relationship between self-esteem scores and frequency of distal-associated ideals or proximal-associated ideals. Qualitative analysis illustrated qualities admired by participants. Implications for clinicians, parents, and educators were discussed.