This article provides the results of a statistical analysis of factors affecting the probability of establishment of a Special Economic Zone (SEZ) in the region, with the use of probit regression. The obtained empirical results indicate that SEZs were established in regions with a high share of industry in employment, with a high rate of unemployment, and at the same time a fast increase in the number of employed persons. This is a confirmation of theses according to which SEZs have become an instrument of provisional restructuring of industry in regions particularly badly affected by unemployment at the start of systemic transition, instead of being a component of a regional policy aimed at providing foundations for a long-term development of economically underdeveloped regions, in accordance with the European Union and World Trade Organisation principles.