Across the world, teacher quality has come to be recognised as one of the most important variables affecting student outcomes; consequently, the regulation of entry into the profession is the subject of iterative review. The traditional 'one-off' interview, involving an interviewee and two or more interviewers, is a common, but not unproblematic, selection mechanism in the field. In particular, the modest positive correlation between performance at interviews and in clinical settings raises questions about using interviews as a selection mechanism for Initial Teacher Education (ITE) programmes. In this paper, we draw on validity theory and some key commentaries and studies in the research literature to offer a perspective on the extent to which the traditional interview provides data that can be used to make good decisions about applicants for ITE. The paper proposes a validity-based framework for use by practitioners to enhance the conceptualisation, design and evaluation of interviews in the process of teacher selection.