Much research has been made in recent years on the effect of over and under education on earnings. In this paper, I examine the validity of the assumption that the return to over (under) education is independent of the level of education. I use data from the PIAAC survey to estimate the return to education. The findings show that while this assumption might be valid for the return to under-education, it seems to be wrong for estimating the return to over-education. I offer an alternative model where the return to over-education or under-education is not linear and depends on the level of education.