We apply a quantile version of the Oaxaca-Blinder decomposition to estimate the counterfactual distribution of the test scores of Black students. In the Early Childhood Longitudinal Study, Kindergarten Class of 1998-1999 (ECLS-K), we find that the gap initially appears only at the top of the distribution of test scores. As children age, however, the gap at the top shrinks whereas the gap in the middle part of the distribution grows. Moreover, the gap due to differences in the effect, rather than amount, of characteristics becomes important. (Contains 3 tables, 2 figures and 9 notes.)