The purpose of this paper is to compare the efficiency of Spanish public and publiclysubsidized private high schools by Data Envelopment Analysis (DEA), employing the results provided by a hierarchical linear model (HLM) applied to PISA-2006 (Programme for International Students Assessment) microdata. The study places special emphasis on the estimation of the determinants of school outcomes, the educational production function being estimated through an HLM that takes into account the nested nature of PISA data. Inefficiencies are then measured through DEA and decomposed into two types: managerial (related to individual performance), and program (related to structural differences between management models), following the approach adopted by Silva Portela & Thanassoulis (2001). Once differences in students’ background, school resources and individual management inefficiencies are removed, the results reveal that Spanish public high schools are more efficient than their publicly-subsidized private equivalents.