Despite the common perception that African-American coaches face discrimination obstacles, only two rigorous statistical studies exist that actually address the issue of racial variation in retention of coaches. Neither study accounts for variation in the level of talent across coaches and the production of wins. We examine the difference between retention of African-American and white NBA coaches based on technical efficiency calculations from stochastic production frontier estimates of team win production. First, we detect no difference in technical efficiency by race of the coach. Second, the evidence is consistent with the idea that coaches are retained based on their technical efficiency. Finally, the evidence fails to support any difference in retention of NBA coaches by race. These results offer finance insights for practitioners and raise methodology questions about retention discrimination findings in the NFL.