This paper argues for an expanded conception of test validity, in which teachers, as end-users of tests, contribute a distinctive perspective on validity, referred to as inferential validity. It also offers a methodology that could be adopted in order to subject this dimension of validity to scrutiny. An investigation conducted into the meanings constructed by teachers of a literacy test, the Emergent Literacy Baseline Assessment (ELBA), is reported to illustrate the methodology. In the first section of the paper, current conceptions of validity are discussed. It is argued that the validation process for tests should include the clarification and justification of the interpretations and uses of observed scores. This argument is illustrated from the methodology for investigating the validity of the ELBA. Self-assessment questionnaires and focus-group interviews provided data on teachers' views about the validity of the ELBA. Arguments in favour of investigating the validity of large-scale tests by taking into account teachers' perspectives are provided.