THE presence or absence of comparative advantage is important to many fields of inquiry besides international trade. Recently, comparative advantage in the performance of tasks by individuals has been shown to be a minimum requirement for the distribution of labor earnings to be different in form from the distribution of abilities (Sattinger, 1975). In particular, skewness in the earnings distribution arises from comparative advantage. While many economists would take the existence of comparative advantage in individuals for granted, much economic analysis and empirical work depends on the absence of comparative advantage. Sherwin Rosen (1977) has recently analyzed the difficulties in labor aggregation introduced by the presence of comparative advantage. Also, the efficiency units assumption requires the absence of comparative advantage in individuals. In turn, the efficiency units assumption is necessary for empirical work relating the distribution of earnings solely to individual characteristics (Sattinger, 1977), as has often been done in the human capital literature. This paper uses data from mechanical aptitude tests (Paterson and Eliot et al., 1930) to investigate comparative advantage in the performance of tasks by individuals. Using both non-parametric and parametric tests, such comparative advantage is found to exist. The paper then presents a method for obtaining a measure of difficulty given a cardinal measure of abilities, and tests for the possibility of intransitive comparative advantage. Artificial economies are then constructed to reveal the effects of comparative advantage on the distribution of earnings. II. Sources of Comparative Advantage