The Mini‐Mental State Examination was modified for use in a non‐western elderly population and named the Korean Version of the Mini‐Mental State Examination (MMSE‐K). Study subjects were a group of normal subjects and a group of patients with functional mental disorders. Among the variables of age, sex, residence, education and diagnosis, education was the only factor found to influence total MMSE‐K scores. Among the component parts of the MMSE‐K, orientation in time, orientation in place, concentration/calculation and language function were significantly influenced by education. After adding one point to scores of orientation in time, two to scores of concentration/calculation and one to scores of language function in non‐educated individuals, differences between total scores and scores of the three items corrected lost statistical significance between the educated and non‐educated elderly groups. Cutoff points for cognitive impairment and their diagnostic validity are not presented in this article. Copyright © 1990 John Wiley & Sons, Ltd.