Adam Smith's idea that the wage structure reflects differences in work conditions is a central tenet of modern competitive theory of the labor market. However, the empirical relevance of this theory of equalizing differences remains unclear. This paper suggests a novel test for compensating wage differentials based on job satisfaction and wages. If wages differentials solely reflect compensation for work conditions, workers will not prefer jobs with high wages to jobs with low wages. Moreover, this new test allows discussing whether industry and firm size wage differentials reflect rents or compensate for work conditions. Results indicate that wage differentials do not exclusively reflect compensation for work conditions.