This article aims to show the segregating effect of the market-like matching of students and schools at the basic school level. The natural experiment case is Tallinn, the capital of Estonia. The current school choice mechanism applied in this case is based on entrance tests. There are increasingly over-subscribed intra-catchment area public schools, where high reputations are reinforced by publicly reported league tables. The current mechanism has resulted in parental strategies of prep-schooling and the manipulation of addresses. Logistic regression results based on survey data assure that, under competitive entrance, families' educational strategies and background characteristics determine the success of admission to schools with good reputations. Understanding this heterogeneous strategic behaviour is important for the effective design of school choice mechanisms.