An adoption/twin design was used to address issues concerning the relationship between education and Mini-Mental State Examination (MMSE) performance. Data on general cognitive abilities, education, and MMSE from 110 identical twin pairs and 177 fraternal pairs aged 50 through 88 were analyzed using multivariate quantitative genetic methods. Genetic influences account for 32% and 19% of the variation in MMSE for men and women, respectively. Furthermore, all of the genetic variation in MMSE is in common with genetic variation for cognitive abilities and education. Finally, correlations between education and MMSE (.21 and .16 for men and women, respectively), which were reduced to near zero when cognitive abilities were partialed out, are primarily attributable to genetic effects for cognitive abilities. These results support the hypothesis that the association between education and MMSE performance predominantly reflects genetically mediated cerebral capacity rather than test bias or health-endangering life styles in less educated individuals.