Using data from NLSY97 we analyze the impact of education on health behavior. Controlling for health knowledge does not influence the impact of education on health behavior, supporting the productive efficiency hypothesis. Although cognition, as measured by test scores, appears to have an effect on the relationship between education and health behavior, this effect disappears once the models control for family fixed effects. Similarly, the impact of education on health behavior is the same between those with and without a learning disability, suggesting that cognition is not likely to be a significant factor in explaining the impact of education on health behavior.