The Dutch educational system has always had a religion-based segmented, ‘pillarized’ structure. The segregation based on ethnicity that has become so conspicuous lately can only to some extent be related to this pillarization directly (namely where Islamic and Hindu schools have been founded). Apparently, factors other than the pillarized structure of the education system alone are required to explain ethnic segregation. How relevant is it that most minorities in the Netherlands have a one-sided, i.e. low, composition in terms of social class? How does segregation based on social class relate to the (formal) segregation based on religious diversity? The analyses in this article are based on data from 550 primary schools in The Netherlands, and show that all major religious segments or pillars within the school system incorporate the complete range of schools segregated on the basis of socio-ethnic class, including black schools. Major features of the curriculum that are indicative of the school’s anticipation of students from privileged or, conversely, deprived population groups appear not to be related to pillarization, but to the social class of the student population. Regarding religion-related curricular differences, some small effects occur.