The Montreal Cognitive Assessment (MoCA) is used to help differentiate normal cognitive aging from Mild Cognitive Impairment and dementia, however, the effects of years of education are only minimally taken into consideration. We have investigated the use of the Spanish version (MoCA-S) in a population with little or no formal education in the Andes Mountains of Colombia, to assess the impact of education on scores. The population was screened for dementia using the Leganes Cognitive Test (LCT) developed for populations with none/few years of education. One hundred and fifty community-dwelling subjects aged 65 to 74 years old were recruited at elderly community social centers. They had a mean of 4.8 (SD ± 3.5) years of education. The MoCA-S and the LCT were administered by trained interviewers. The association between the MoCA-S total score and its sub-scores by cognitive domain, and the years of education were examined by ANOVA in the 150 subjects and in the “free of dementia” subjects (n=126) group. Twenty-four persons (16%) tested positive on LCT screening for dementia. MoCA-S scores varied significantly depending on education in the whole sample and in those without dementia. In the whole population, including those who screened positive for dementia, the mean MoCA score was 17.61 (±4.9). By educational achievement, MoCA scores were 16.16 (±4.34) for those without completing primary school or illiterate (n=74), 18.28 (±4.53) among those with primary school (n=46), and 20.13 (±5.82) among those with more than primary school (n=30). There were significant differences by education in the total score (P<0.001) and the subscores for executive function-visuospatial tasks (P<0.008), language fluency (P<0.007), verbal abstraction (P<0.001), and serial substraction (P<0.001). Similar results were found in the “no-dementia” group. Orientation, delayed recall and language (repeat sentences) were less influenced by education. Scores from the MoCA-S test are strongly dependent on education achievement among elderly Colombians with no or few years of education. Very low educational achievement and illiteracy must be considered when interpreting the MoCA scores given the high error rate on items that depend on culture and the ability to read and write.