This study uses US survey data and Swedish register data to estimate and compare the relationship between returns to schooling and ability. A significant and positive relationship is found for Sweden, but not for the USA. Based on the predictions of the optimal schooling model it is argues that measured differences in the relationship between returns to schooling and ability could depend upon on differences in the schooling systems between the countries. The findings suggest that a low price of higher education in Sweden makes a relationship between returns to schooling and ability observable in Sweden but not in the USA.