This paper explores reasons why secondary schools with a religious character have pupil intakes that are of a higher social background and ability than their secular counterparts. We show that this is especially true across all regions in England once the characteristics of the pupils living in the local neighbourhoods are taken into account. Data from the National Pupil Database and the Longitudinal Survey of Young People in England are combined to show that likely reasons for this are complex. Parents reporting a religious affiliation are more likely to be better educated, have a higher occupational class and a higher household income. We also show that higher-income religious families are more likely to have a child at a faith school than lower-income religious families. Policy implications regarding the state-funding of faith schools are discussed. (Contains 7 notes, 8 tables, and 4 figures.)