The Mini-Mental State Evaluation (MMSE) was used in an epidemiological survey of a community of mixed ethnicity (Hispanic, white non-Hispanic) as part of the Los Angeles Epidemiologic Catchment Area Program. Results of the study showed that age, educational level, ethnicity, and language of the interview influenced the number of MMSE errors. Items on which the effects of ethnicity and language were most pronounced were identified, and suggestions on ways to minimize such sociocultural artifacts are provided in efforts to improve the epidemiological significance of the instrument, particularly as it concerns cross-cultural research.