In this paper we evaluate the Swedish Establishment Reform, carried out in 2010 with the goal of speeding up the establishment of refugees and their family. From December 1, 2010 the reform transferred the responsibility for the integration of newly-arrived refugees from the municipalities to the government funded Public Employment Service through which those eligible should get establishment talks, individual plans and coaches. The Reform was motivated by concern over the low employment level and slow integration of refugees. Our approach is to compare the outcomes of the Treatment group, which took part in establishment activities and arrived between December 1, 2010 and December 31, 2011, to those of the Comparison group, which arrived in the eleven months preceding the Reform and participated in municipal introduction programs, controlling for a rich set of observables, including country of birth and month of residence permit. Outcomes are measured in terms of employment and earnings in 2012, 2013 and 2014 for the Treatment group and in 2011, 2012 and 2013 for the Comparison group. Our data comes from registers held by Statistics Sweden and covers all immigrants. The results suggest positive and significant effects of the Reform. In the second year after program-start the Treatment group had about 5.7 percent higher probability of employment and in the third year about 7.5 percent higher. The effects on earnings were larger, about 20 percent higher earnings for the Treatment group after the second year and about 22 percent higher after the third year.