The study investigates the mean difference in reading achievement between third-graders in public and independent schools in Sweden. The data come from the Swedish participation in PIRLS 2001 conducted by IEA. Variables from the home questionnaire mainly indicating possession of cultural capital are used as independent variables. A total IRT score of reading achievement is used as outcome variable. It is shown that students in independent schools have on average better reading results and also a more advantageous socio-economic background than have students in public schools. Social selection hence characterises independent schools. In addition, indicators of students' cultural capital account for the major part of the mean difference between school types. (Contains 6 tables and 2 figures.)