Background: Educational differentiation, or the allocation of students to different types of school or different locations within schools, is a common feature of modern education systems. In most countries, student outcomes vary considerably, both by school and academic location within schools. Purpose: The purpose is to ascertain the extent that between- and within-school differences in student performance can be attributed to students' socio-economic background and their home resources. Sample: The samples are from the OECD's 2000 Program for International Student Assessment (PISA) study of 15-year-old students' performance in reading, mathematics and science. Data from 30 countries were analysed. The student samples ranged between 2500 (USA) and 30,000 (Canada) students. The samples are representative of the populations of 15-year-old school students in each country. Design and methods: Multiple regression analysis was used to ascertain the extent to which between-school differences (measured by intra-class correlations) and within-school differences (measured by the effects of the composite variable "academic location" which comprised grade and school programme) declined when taking into consideration students' socio-economic background and home environment. Results: In most countries the intra-class correlations and the effects of academic location declined only marginally. This was especially the case in countries with large between-school differences such as the tracked systems of continental Europe. Conclusions: Differences in student performance between and within schools cannot be accounted for by socio-economic background. The study lends some support to the official rational for tracking and other forms of educational differentiation, that the allocation of students to different school types and academic locations within schools is largely based on student ability.