The volume of work published by accounting scholars on corporate social responsibility (CSR) reporting is impressive. As a community we have explored ongoing developments in the field, including the accuracy of the accounts and their economic impact for various actors. We have also asked broader questions about the effect of CSR reporting on corporate accountability and responsibility. However, as it currently stands, our interests are separated because of specific epistemological approaches, and these have divided our efforts. Drawing inspiration from Donaldson and Preston (1995), we identify three streams of CSR reporting research, that is, descriptive, instrumental, and normative. Descriptive research asks: what is CSR reporting? It explores the content of reports and the effect of various contexts on the practice and quality of reporting. Instrumental research asks whether CSR pays, and explores the commercial and economic benefits of CSR for firms and the information value of disclosures for investors. Normative research examines the significance of CSR reporting at a societal level, asking whether these practices have improved organizations’ accountability to their stakeholders. While the main purpose of our paper is to review the CSR reporting literature (245 articles), we also map some of the positive impacts that may result from greater epistemological engagement between streams. We identify specific aspects of each stream that might benefit a cross‐pollination of ideas, findings, and insights. We argue that future CSR reporting will be enriched by conversations that extend beyond epistemological divides, which are more likely to mirror the complex reality of CSR in practice.