Many introductory research methods textbooks still divide the research paradigms into two broad approaches-positivist/quantitative/deductive and interpretive/qualitative/inductive. However, this bifurcation of research orientation does not do justice to the philosophical and methodological pluralism present within business research. This paper offers a third way by reflecting on a retroductive analysis in a critical realist case study. The philosophy of critical realism employs retroductive analysis to search for mechanisms underpinning the empirically observed events. Mechanism-based theorising is a suggested way in the business research to develop middle-range 'sometimes true' theories. This paper demonstrates the process of retroduction for the identification of mechanisms through an illustration of the data collection, coding and analysis process in a multilevel critical realist case study. In the process, it outlines the challenges faced and offers suggestions to overcome those challenges. This paper does not claim to provide a set of best practices for multilevel retroductive analysis. However, it is hoped that it sensitises business researchers to explore the critical realist perspective and to employ retroduction for mechanism-based theorising. © ACPIL.