This study tests the hypothesis that a disjuncture between an individual's attained level of education and that held by average workers in the individual's occupation leads to higher mortality among those with a prolonged mismatched status. Swedish register data are used in a 19‐year longitudinal mortality follow‐up study of all causes and specific causes of mortality. Participants were all men and women born between 1926 and 1985 who were alive on 1 September 1990, who had concurrent information on their attained level of education and the specific occupation or industry they were employed in during this period for at least a consecutive year. An objective measure of educational and occupational mismatch was constructed from these data. Those with a stable, over‐educated matched, or under‐educated employment status are included in the final analysis (N = 2,482,696). Independent of social, family, employers’ characteristics and prior health problems, the findings from a multivariate, stratified Cox regression analysis suggest there is excessive mortality among the over‐educated, and a protective effect of under‐education among native‐born Swedish men and women.