In Spain, the number of job mismatch graduates is higher than in other surrounding countries. The reasons for this phenomenon differ, to a certain extent, according to the gender of the worker. This research analyses the educational factors that best predict education-job match in men and women. The two most frequent education-job match types have been analyzed: vertical match and horizontal match. Data have been obtained from the 2014 University Graduates Labor Insertion Survey prepared by the National Statistics Institute. We have worked with a sample of 19272 individuals. In addition to contingency tests, the predictive value of the educational variables was explored using binary logistic regression analysis. We find, as a main result, that variables such as the area of study, the possession of another university degree, the level of languages or the knowledge of new technologies have a greater influence on the education-job match in the case of women. The results can be explained on the basis of different theories which indicate that the reasons that lead an individual to choose a job not related to his training differ according to the gender of the worker. Family obligations continue to be elements that condition, to a large extent, women's employment decisions. Labor market seems to value the skills acquired by men and women differently.