We investigate the impact of competition between primary schools on the quality of education. Do schools facing more competition in their neighbourhood perform better than schools facing less competition? As a main measure of school quality, we look at the performance of pupils at the Cito-test. Due to potential endogenous school location and pupil sorting, we adopt an instrumental variable strategy (IV) in the spirit of Gibbons et al. (2008). Using a large range of data on pupil, school and market characteristics, our results suggest that school competition has a positive effect on pupil achievement measured by the Cito-test.