Objective: Prior research (Manley et al., 2004) has suggested that years of education may not be an adequate predictor of neuropsychological performance especially amongst African Americans. Quality of education, as defined as performance on the Wide Range Achievement Test-3, Reading subtest (WRAT-3), was found in fact to be a better predictor of cognitive functioning in their sample of healthy community dwelling adults. The current study sought to expand on this work by examining this hypothesis in a clinical population using updated measures. Specifically, the current study examined whether educational quality was a better predictor of cognitive performance in a Midwestern clinical sample of elders than educational quantity. Method: Retrospectively, neurocognitive data was gleaned from the files of 60 consecutive African American (mean age = 77.1, SD = 7.2; mean education = 14.7, SD = 3.3) and 60 White (mean age = 79, SD = 6.4; mean education = 13.5, SD = 3.5) patients. Simple regression models determined the relationships among each of five predictor variables (age, sex, years of education, race and WRAT-4 reading) and performance on neuropsychological measures. Results: Significant correlations were documented between WRAT-4 reading and race, (p < .001) and years of education (p < .01). Reading ability significantly predicted performance on neuropsychological measures above and beyond age, gender, and years of education, across racial groups. Conclusion(s): Quality of education was found to be a better predictor of cognition regardless of race. Implications for test interpretation among diverse older adult patients will be discussed.