Overeducation can be measured through several approaches, each one leading to different results as to its incidence. This paper aims to build a new and up-to-date measure of overeducation, the Skill Analysis measure, based on the content of jobs in terms of tasks and skills required to perform a job. To this end, I mobilise data from the Occupational Information Network (O*NET) and information about learning outcomes from the European Qualifications Framework. I then compare the new measure of overeducation to usual statistical measures using the French Labour Force Survey. Results confirm that objective and statistical measures do not apprehend the exact same matter, showing only moderate correlations between approaches.