The main aim of this paper is to examine the effect of the educational institution, and, specifically, its public or private ownership on the results of the Spanish education system. The empirical analysis is based on the application of multilevel regression techniques on data from OECD PISA-2003. The results of the analysis support the working hypothesis, in which it is stated that the best performance of private institution pupils are not explained by ownership, but rather by variables such as the individual and family characteristics of the users, their peers, and other variables related to the institution.