Internationally, quality assurance schemes persist despite long-standing dissatisfaction and critique of their impact and outcomes. Adopting a critical systems perspective, the article explores the relationships between the knowledge, power and meanings that stakeholder groups bring to the design and implementation of quality assurance systems. The analysis shows that such systems are designed to serve the external accountability purposes of government and agencies outside the university who are responsible for designing the systems. Academics inside the university are affected by quality assurance systems but uninvolved in their design. The knowledge and power distance and differences of meaning between the system designers and academics result in quality assurance systems that are unable to contribute to the improvement of teaching and research in the university. The article proposes interconnected but clearly differentiated definitions of quality assurance and quality improvement that can inform systems design aimed at more than meeting external accountability demands.