This study examines the proposition that secondary school choice in England has produced a stratified education system, compared with a counterfactual world where pupils are allocated into schools based strictly on proximity via a simulation that exploits the availability of pupil postcodes in the National Pupil Database. The study finds current levels of sorting in the English secondary school system-defined as pupils who do not attend their proximity allocation school-to be around 50 per cent, but estimates that only one-in-five pupils are potentially active in sorting between non-faith comprehensive schools. School segregation is almost always lower in the proximity counterfactual than in the actual data, confirming that where pupils are sorting themselves into a non-proximity school, it does tend to increase social and ability segregation. The difference between school and residential segregation is greatest in urban areas and LEAs with many pupils in grammar and voluntary-aided schools.