Two vital American social institutions, education and health care, are intimately related. The individual-level relationship between education and health is tenuous, changing over time, throughout the life course, and across generations. Previous research in this area does not separate the mediating effects of age, period, and cohort or assess the unique effects of various levels of educational attainment on health. Using longitudinal data from the Panel Study of Income Dynamics, this dissertation untangles these factors to find that education has become a more important predictor of health now than twenty years ago, education benefits health across the life course, and college education is necessary for good health in recent generations whereas high school education was a more important predictor of health in earlier generations. From a theoretical standpoint, this study illustrates the need for a more prominent theory to explain the changing nature of the education and health relationship. Methodologically, results suggest that longitudinal analysis is a superior technique to cross-sectional analysis, as the effects of education on health are suppressed in cross-sectional analyses. From a policy standpoint, findings indicate that one viable solution to decreasing health disparities is improved access to education, rather than improved access to physicians, which is the dominant solution in today’s society.