How does the relationship between earnings and schooling change with the introduction of comprehensive economic reform? This paper sheds light on this question using a unique data set and procedure to reduce sample selection bias. Our evidence is from consistently coded, non-retrospective data for about 4 million Hungarian wage earners. We find that returns to skill increased by 75 percent from 1986 to 2004 (that is, during the period stretching from communism to full membership in the European Union). Moreover, our results identify winners and losers from reform. Winners were the college and university educated and those employed in the services sector (which excludes those in public services). Our results show that reform losers were those in construction and agriculture, those who attained only primary or vocational education (who actually experience a decrease in the returns to their education) as well as those younger workers which acquired most of their education after the collapse of communism (that is, after the main reforms were in place).