'Personalised learning' and the value of national assessment data in achieving it have been identified by the UK Secretary of State for Education and Skills as essential for raising educational standards. Employing multilevel analysis, this paper compares children's end of primary school (Key Stage 2) test scores with those they achieved in comparable test papers taken in each term of their first year of secondary school. The paper questions the reliability of national assessment data in respect of the performance of individual children, their predictive validity and thus their value in contributing to the provision of 'personalised learning'.