Business process change involving information technology has become notorious for high failure rates. Apparently, the larger the project scope and size, the higher the failure rate. Also, as a major factor among others associated with failure, past research has identified an overemphasis on technical validity and functionality at the expense of the social and institutional process side in change projects. Yet, even if important factors had been correctly and completely identified, the social process, within which they interact, might have remained obscured. Furthermore, long-term, large-scale change projects, particularly when they encompass IT components as major building blocks, defy the conventional start-to-closeout, detailed project planning. Rather, they unfold in a stepwise fashion with alternative paths towards a coarse goal, particularly in environments with distributed powers, as in the public sector. For this process and this environment, the author proposes that participatory action research (PAR) has the capacity to provide a holistic methodological and social-process frame and a mode of joint inquiry. This article gives an account of the approach based on a large government change project in its early stages. In this project, practitioners and researchers formed a joint team, which strove to lay foundations to a multi-year, IT-enabled change initiative meant to become a state-wide exemplar for similar projects. During this early project stage, salient stakeholders and their needs were identified, which helped launch a highly inclusive process. However, as found in this project, external influences impacted this public-sector project in a way that the build-up of consensus remained incomplete, underscoring the need for ongoing stakeholder involvement within the project.