Topics are often used as a key speech elicitation method in performance-based assessments of spoken language and yet, the validity and fairness issues surrounding topics are surprisingly under-researched. Are different topics 'equivalent' or 'parallel'? Can some topics bias against or favour individuals or groups of individuals? Does background knowledge of topics have an impact on performance? Might the content of test taker speech affect their scores - and perhaps more importantly, should it? This monograph draws on original data as well as insights from empirical and theoretical research to address these questions. Grounded in the real-world assessment context of IELTS, this volume explores issues related to topic validity against the backdrop of one of the world's most high stakes language tests.