We examined trends in the incidence and correlates of educational and skill mismatch in the United States. We focused on trends over time in the associations between various types of mismatch and a range of factors including contextual conditions. We explored whether contextual conditions at the transitional period from school to jobs increase or decrease the probability of mismatch and whether such relationships persist throughout the working career. Our central questions were how the incidence of and relationship between educational and skill mismatch in the U.S. changed between 1994, 2003, and 2012 and how this differed by age, gender, immigration status, educational attainment, and occupation. We used three cross-sectional surveys that had not previously been implemented for such an effort. These were the International Adult Literacy Survey (IALS) in 1994, the Adult Literacy and Life-skills (ALL) survey in 2003, and the Program for the International Assessment of Adult Competencies (PIAAC) in 2012. Repeated cross-sectional data provided us with substantial analytic leverage. Our findings point toward the key role of occupational or positional factors rather than individual worker characteristics as being most implicated in trends in mismatch. We describe the importance of our results for labor market theories.