In recent years enterprise integration has become an important theme in information systems research and practice. Enterprise integration pervades enterprise resource planning (ERP), supply chain management (SCM), and customer resource management (CRM) systems and applications. However, much IS research in this area seems to assume that enterprise integration is inevitable and that enterprise integration is strategically necessary. These assumptions appear to have held sway until now, probably because most IS researchers have studied the life cycle of just one single enterprise systems project. Our research throws into question these assumptions. Using critical ethnography, we studied a small-to- medium sized company within the context of a large conglomerate over a seven-year period in total, from mid 1996 to early 2004. This company was one of many subsidiaries within a large conglomerate in the Asia- Pacific region (one of the global 500 companies with annual revenues of more than US$4 billion). Looking at the broader context (the conglomerate as a whole), and seen over a sufficiently long time scale, our findings suggest that enterprise integration is not inevitable, nor is it always strategically necessary. Rather, enterprise integration is perhaps best described as a cycle: as one or more cycles of integration, disintegration, and (perhaps) reintegration. This paper can be seen as one response to the call for more empirical in-depth case studies concerning enterprise systems.