The issue of racial inequality in education has consistently been addressed through government policy in an attempt to solve the problem of discrimination in the American school system. The latest government attempt is the No Child Left Behind Act of 2001 (NCLB). In order to examine the effectiveness of the NCLB, this study examines the relationship between race and composite reading and math test scores. Secondary data analysis was used from the Educational Longitudinal Study of 2002 (ELS:2002) of 8,215 10th grade students with a composite model consisting of: student role performance (SRP), schools, families, and peers. Univariate, bivariate, and multivariate analyses are used to examine the independent effects on test scores. Mean composite test scores show an 82.8% gap in test scores between black (44.42) and white (53.64) non-Hispanic 10th grade students. The examination of the unique variance in test scores for each model segment shows student role performance factors and family factors explain more of the “unshared” variance in test scores than other model segments. These results suggest that discrimination based on the ability of a student to perform the roles associated with their position within the school setting as well as family factors outside the school setting contribute to the academic achievement gap between blacks and whites.