Like many other countries in Central and Eastern Europe, children in Slovakia are allocated to different types of schools at an early age based upon their perceived aptitude. Part of the selection process includes an attempt to identify those children who are particularly academic-oriented. Primary and secondary education in Slovakia is divided into two separate phases: lower and upper, which are also reflected by educational levels two and three, set by the International Standard Classification of Education (ISCED). Primary and lower secondary schools represent the compulsory phase of the educational path and the institutional structure through which all children pass. The upper secondary schools belong to the post-compulsory phase of school attendance for young people who want to enter the labour market or continue their studies. Primary or basic school ("zakladna skola") takes 9 years to complete, and provides primary education (Grades 1-4) and lower secondary education (Grades 5-9). At the end of Grade 9 students are streamed into several types of upper secondary school. There are two branches of upper secondary schools in Slovakia: grammar school or high school ("gymnazium"), and various vocational schools ("stredna odborna skola," "stredna odborna uciliste," "zdruzena stredna skola"). In this paper, the authors' focus is primarily on the selection of students into "gymnazium" schools prior to grade 9, and in the transition from grade 9 to grade 10, and in the differences in results and access to resources in the two sectors. Although in the second stage of selection students are also selected into two types of "non-gymnazium" secondary schools, they are primarily concerned with the differences between "gymnazium" and "non-gymnazium" schools. In April 2003, 15-year old students from 281 schools in Slovakia took part in the 2nd cycle of the OECD PISA study. This paper aims to examine the data from the enhanced sample of the PISA 2003 study to: (1) examine the extent to which early and late selection are related to family socio-economic status and gender; and (2) determine the extent of inequalities in mathematics achievement among students in the two types of schools, before and after controlling for students' socio-economic backgrounds and the SES composition of the school. (Contains 3 figures and 3 tables.)