Rapid changes in data forms and availability and evolving requirements in external quality assurance regimes are combining to prompt reconsideration of the nature of evidence in quality assurance. Although evidence is a longstanding feature of both internal and external quality assurance, much of the commentary is limited to examples of types of evidence with less attention paid to what constitutes good evidence. Evidence for quality assurance can be categorised as being pre-existing or bespoke, qualitative or quantitative, and for and from particular groups or perspectives. Each of these categorisations has implications for how evidence should be considered by universities in their self-review activities and by external audit panels. The quality of evidence itself can be assessed using criteria of relevance, representativeness, verifiability, its cumulative nature, whether it is actionable, contextual and holistic and able to be triangulated. The academic audit framework for the sixth cycle of academic audit for New Zealand universities applies to all students, all delivery and all staff who teach or supervise or support teaching or supervision. Consideration will need to be given to how evidence reflects this embedded or systemic approach to academic quality. The paper develops guidelines for reconsidering evidence in quality assurance.