Since cutoff scores are necessary for differentiating normal cognitive functioning from dementia, the main objective of this study was to establish normative data of the K-T test, a standardized cancellation test. The construct validity of K-T test was also investigated. In total, 2471 cognitively intact elderly subjects from the SU.VI.MAX 2 study were included in the present study. Younger subjects, women and subjects with higher education had a higher number of correct answers and fewer omission errors. Participants made few commission errors and only the educational level was found to have a significant effect. Normative data for correct responses and efficiency were stratified by age, education, and gender. Correlation analysis showed that the K-T test was significantly correlated to validated neuropsychological tests assessing executive functions. This study permits to strengthen the utility of the K-T test to detect impairment of the executive components involved in the task.