This paper presents new evidence on the relationship between exposure to an own-race teacher and student achievement, using data from Tennessee's Project STAR class-size experiment. The experiment randomly assigned students and teachers within participating schools, allowing for a more accurate assessment of the effects of racial pairings. The results indicate that a one-year assignment to an own-race teacher significantly increased the math and reading achievement of both black and white students by roughly three to four percentile points. The study also finds that the racial pairings of students and teachers were independently assigned, and that the within-school variation in exposure to an own-race teacher is uncorrelated with other important student traits. The results suggest that exposure to an own-race teacher generates substantive gains in student achievement for both black and white students. However, the estimated effects of assignment to an own-race teacher appear to vary with respect to other student, teacher, and classroom characteristics. The study also addresses concerns about the validity of the experimental design, including treatment crossover and attrition, and finds that these issues are not problematic in this context. Overall, the results provide strong evidence that exposure to an own-race teacher is associated with higher student achievement, and that the randomized pairings in the Project STAR experiment provide a unique opportunity to evaluate the educational impact of own-race teachers. The study also highlights the importance of considering the role of teacher quality and the potential for unobserved teacher characteristics to influence student achievement. The results suggest that the educational benefits of own-race teachers may be driven by passive teacher effects, such as role-model effects, rather than active teacher biases. The study concludes that the findings have important implications for educational policy, particularly in the recruitment and retention of minority teachers.This paper presents new evidence on the relationship between exposure to an own-race teacher and student achievement, using data from Tennessee's Project STAR class-size experiment. The experiment randomly assigned students and teachers within participating schools, allowing for a more accurate assessment of the effects of racial pairings. The results indicate that a one-year assignment to an own-race teacher significantly increased the math and reading achievement of both black and white students by roughly three to four percentile points. The study also finds that the racial pairings of students and teachers were independently assigned, and that the within-school variation in exposure to an own-race teacher is uncorrelated with other important student traits. The results suggest that exposure to an own-race teacher generates substantive gains in student achievement for both black and white students. However, the estimated effects of assignment to an own-race teacher appear to vary with respect to other student, teacher, and classroom characteristics. The study also addresses concerns about the validity of the experimental design, including treatment crossover and attrition, and finds that these issues are not problematic in this context. Overall, the results provide strong evidence that exposure to an own-race teacher is associated with higher student achievement, and that the randomized pairings in the Project STAR experiment provide a unique opportunity to evaluate the educational impact of own-race teachers. The study also highlights the importance of considering the role of teacher quality and the potential for unobserved teacher characteristics to influence student achievement. The results suggest that the educational benefits of own-race teachers may be driven by passive teacher effects, such as role-model effects, rather than active teacher biases. The study concludes that the findings have important implications for educational policy, particularly in the recruitment and retention of minority teachers.