This paper employs recently developed econometric models of marginal treatment effects to analyze the relevance of labor market comparative advantage and segmentation in the participation and earnings performance of workers in formal and informal jobs in Argentina. A novel household data set on informality and self-employment and information on labor inspections targeting informal work was collected for this purpose. We account for endogeneity and selectivity issues in our estimations. Our results offer evidence for both comparative advantage and segmentation. No significant differences between the earnings of formal salaried workers and the self-employed are found, once accounted for positive selection bias into formal work. This is consistent with labor market comparative advantage considerations. On the contrary, informal salaried employment carries significant earnings penalties, alongside negative selection bias and modest positive sorting. These results are more consistent with segmentation.