Objective: The use of educational attainment as a demographic correction for neuropsychological performance among minorities may be inadequate, due to disparities in quality of education between racial/ethnic groups. As such, reading level, which may better reflect education quality, may be a more meaningful measure. Here, we evaluated the relationship of reading level, years of education, and a CD4+ T-cell nadir with global neuropsychological performance in an urban HIV-infected cohort. Method: 50 HIV+ , predominantly African-American participants (27M, 23F), ages 23–71 (M= 42.18 ± 10.95) were recruited through the Comprehensive NeuroAIDS Center at Temple University. Hierarchical multiple regression analyses assessed the predictive abilities of reading level, as measured by the Word Reading subtest of the Wide Range Achievement Test-4 (WRAT-4), years of education, and CD4+ T-cell nadirs on combined measures of cognitive performance [Global Cognitive Index (GCI)]. Results: Years of education and CD4+ T-cell nadirs reflected 2.6% of the variability in the model (p = 0.544). Inclusion of reading ability (WRAT-4), F (3,46) = 10.824, p< .001 increased the predictive capacity of the model. This model demonstrates reading ability reflects 38.8% of GCI variance and predicts that +1 unit in WRAT-4 leads to +0.436 ± 0.079 in GCI. Conclusion: Here, we report that CD4+ T-cell nadirs and years of schooling have little predictive value in determining GCI. In contrast, inclusion of WRAT-4 predicts GCI by approximately 39%. Together, these data demonstrate that the variability in neuropsychological performance is not a result of disease burden but rather poor quality of education and the standard for using years of education as a demographic correction among racial/ethnic minorities may overpathologize these groups.