Three arguments regarding racial equity have arisen in the school choice debate. Choice advocates charge that choice will improve access to quality schools for disadvantaged minority students (Chubb & Moe 1990; Coons & Sugarman, 1978; Godwin & Kemerer, 2002; Viteritti, 1999). Critics argue that choice is unlikely to benefit minority students, but they are divided as to why this may be the case. Some maintain that unfettered choice leads to racial segregation (Henig, 1996; Mickelson, 2005; Saporito, 2003); others maintain that while choice may be successful in reducing segregation at the building level, choice programs may be problematic to the extent that they segregate schools at the classroom level (Wells & Crain, 1997; Wells, Holme, & Vasudeva, 2000; Wells & Roda, 2009; West, 1994). I use data from the eighth-grade wave of the Early Childhood Longitudinal Study to examine these questions. Results indicate that the racial composition of magnet schools is not statistically different from regular public schools; however, magnet schools are more heterogeneous at the classroom level, but only with respect to White/Hispanic racial composition. In particular, honors classes in magnet schools are significantly more diverse than honors classes in regular public schools, but only with regard to White/Hispanic diversity.