<p>Action research (AR) is a process of creating knowledge in and through action, generating outcomes for both practice and theory and is therefore widely used to link evidence and practice, yet data collection, analysis and interpretation within AR cycles are often underreported and poorly understood. The dual purpose of taking action and contributing to theory introduces an inherent tension, often articulated through the interplay of a core action cycle and a thesis or knowledge-generation cycle. While the iterative, participatory nature of action research is well established, the methodological processes through which data are generated, handled, and transformed into actionable knowledge remain comparatively underexamined. Working within the organisation development tradition of action research, this article critically explores how data produced within action research, emerging through cycles of constructing, planning, acting, and evaluating are collected, analysed, interpreted, and integrated across evolving cycles of inquiry. The discussion highlights the need to move “beyond the cycle” towards a reflexive and participatory approach to data practices. It argues that an essential part of this movement begins by continuously aligning the processes of data collection, analysis and interpretation with the Quality Action Research Checklist (QuARC) domains of context, relationships, process and dual outcomes. Such continuous engagement involves questioning how personal assumptions, values, and actions shape the data and its interpretation, and how these influences evolve over time. Hence, rethinking data in AR as a dynamic, co-constructed and theory-building process strengthens its methodological quality and ensures sustainable contributions to knowledge, professional learning and organisational change in many organisational and healthcare care contexts.</p>