We test for the signalling hypothesis versus human capital theory using the Wiles test (1974) in a country which has experienced a dramatic increase in the supply of skills. For this purpose, we construct a job match index based on the usefulness of the school-provided skills and the relevance of the job performed to the field of study. Then we regress the first earnings of graduates on this index using OLS and Heckit to control for omitted heterogeneity of the employed. The data we use come from a representative tracer survey of Poles who left secondary schools or graduated from HEIs over the period of 1998-2005. We find that only the HEI graduates obtain a wage premium from skills acquired in the course of formal education. This finding is robust to a large number of robustness checks with different indicators of the educational mismatch and instrumental variables.