The authors use a switching regression method to test the hypothesis that segmentation in the Chilean labor market exists. The results show that any a priori definition of the sample will lead to rejecting the segmentation hypothesis. (DSE/DÜI)