The principles of new public management-market mechanisms, accountability and standards-have been applied in the education system. These methods are supposed to increase efficiency, but there is also a risk of negative consequences from the services provided if the measures of performance target a reduced range of goals, ignore relevant variables or are not valid measures. Indicators used to compare schools' performance are aggregate measures, such as the percentage of students who have access to secondary education and the average qualification value. This study reports how accountability policy and procedures may affect the functioning of the education system through the case study of a school serving a diverse student population. The school organisation was influenced by measures of performance, external events and contextual and selection variables. The average qualification value measure seems to be a limited measure of performance at the school level, since it largely reflects school composition and school segregation. Even the available performance measures adjusted for background variables do not take account of relevant variables that may influence the school's need of resources and its results, such as students' language proficiency and special educational needs. Other performances that are not easily measured-such as the prevention of dropout, improvement of school attendance and provision of an equitable education for all studentsare disregarded. Schools serving those students with the most needs risk being penalised by an approximate and restricted range of accountability systems because there is a risk that the schools will appear to be failing when they are working with more complex and advanced tasks than average schools. Based on these inaccurate performance measurements, the school may be targeted with wideranging, severe and basically unjust interventions.