The fact that Black student-athletes are awarded more than half of all college basketball playing opportunities and scholarships at NCAA Division I schools suggests that intercollegiate sports may present part of American society free from racial discrimination. However, an entirely different and disturbing picture is presented when one views the racial make-up of college coaches (Hill & Murry, 1997). For example, in Division I basketball, Blacks make up greater than 50% of the student athletes, but only 25% of the head coaches (Earle, 2000). The present study took steps toward understanding to what extent work place diversity has led to such discrepancies. Specifically, the purpose of this study was to test the impact of relational diversity in a coaching staff on three important career outcomes of Division I assistant basketball coaches. We used leader-member exchange (LMX) theory (Graen, Novak, & Sommerkamp, 1982) and the emerging concept of relational demography (Tsui & O'Reilly, 1989) to develop hypotheses regarding the career satisfaction and supervisor mentoring received in coaching. It was hypothesized that being in a high LMX group would result in more positive career outcomes and that these categorizations would be based on the demographic make-up of the dyad (e.g., white head coach-black assistant coach) or coaching staff (e.g., mostly white, mostly black) (Tsui, Xin, & Egan, 1995). A mailed survey design was used to complete the study. Questionnaires using established scales (Greenhuas, Parasuraman, & Wormley, 1990; Schriesheim, Neider, Scandura, & Tepper, 1992) were distributed to 640 Division I assistant basketball coaches and achieved a 37% response rate (N=235, n=74 Black, n=161 White). Analysis of variance was used to assess differences in supervisor support and career satisfaction on the trichotomized LMX variable (low, medium, high). A chi-square analysis was used to assess if the categorizations into an LMX group were a result of dyad dissimilarity or diversity in the coaching staff. Results of the analysis of variance indicated that being in a high LMX relationship was critical toward experiencing greater supervisor support (F=85.745, p < .001) and career satisfaction (F=22.58, p < .001). However, results of the chi-square failed to support the hypotheses regarding diversity in coaching dyads and groups, as dyad race and staff composition did not associate significantly with any of the three LMX categorizations. The findings both contradict and reaffirm the research in this area and explanations using both social psychology and relational demography literature are offered.