In 2001, the Rwandan government began the phased introduction of a decentralisation programme throughout the country. The new programme aimed at countering citizens’ social, economic and political marginalisation which was widely viewed as constituting one of the principal drivers of the horrific genocide in 1994 in which almost one million Rwandans were brutally massacred. This research aims at analysing the extent to which Rwanda’s decentralisation process is meeting these aims. Employing a framework which differentiates between procedural participation (electoral participation); substantive participation (ongoing active participation in local decision making); and participation as cost-sharing (ongoing participation through financial contributions and voluntary labour), four main findings are discussed. First, although there is much talk among officials and commentators about bottom-up planning processes emanating from local village meetings (formerly ubudehe and now umuganda), and while such an ethos underpins the original Decentralisation Policy formulated in 2001, the accompanying legislation is somewhat scant in its references to such a form of participation as well as to mechanisms whereby this might take place. Instead, a focus on community mobilisation in participation as cost-sharing is more apparent. Second, a shift in emphasis within the decentralisation programme over time is evident. The current national strategy of fast-track economic development as a route out of poverty has been superimposed on the original goal of reconciliation and community building with an attendant emphasis on results over process. This is evident in the 2013 revisions to the Decentralisation Policy as well as in the shift, during the second phase, from political decentralisation toward administrative decentralisation evidenced in the introduction of public management frameworks such as the imihigo and the emphasis on the administrative capacity of local leaders. It is also evident in the demise of the local planning function of ubudehe which is now a social categorisation mechanism with final categories being decided by cell leaders. This has necessitated in a shift in emphasis from substantive participation to participation as cost-sharing. Third, the findings from a) the comparison of local official and community priorities, and b) citizens’ knowledge and use of local structures reveal no evidence of representation or accountability at district level where plans and policies appear heavily influenced by national prerogatives and where senior political figures are, paradoxically, both elected through the official system and strategically selected at national level. These same findings reveal some evidence of responsiveness at sector and cell level however, where communications and contact with communities are more frequent. Although more aware of community priority issues and needs, local officials are constrained in their capacity to address these however, due to pressures to meet the ambitious targets set out in their imihigos which draw from higher level plans and targets. And fourth, of the three forms of participation examined, participation as cost-sharing emerges as the most common, with increasing emphasis placed on this in recent years as local entities are encouraged to move toward fiscal autonomy and self-reliance. The heavy emphasis on this form of participation is viewed as problematic in a number of respects. Reflecting on these findings in the context of ongoing debates on the efficacy of supports to the process, three broad questions are posed. First, given the parallels in international views of Rwanda pre-1994 and Rwanda today, how “good enough” does “good enough governance” need to be? Second, is “fast-track development” compatible with other peace-building objectives aimed at transforming the political space resulting in equity, social cohesion and local political legitimacy when the pressures of such a fast-track approach are dependent on high levels of cost-sharing by citizens? And third, when placing capacity and capacity building at the heart of supports to the process, what and whose capacity do we mean? Is it the capacity of communities to substantively interact with detailed administrative policy and budgetary mechanisms within a complex system or is it the capacity of the system to engage with community members as equals, valuing their knowledge and analysis, and developing the capacity, skills and tools to translate this into the necessary policy and budgetary frameworks? The report concludes with a reminder that decentralisation was not just the key mechanism through which communities were physically mobilised during the genocide, it was also (together with the aid industry more broadly), for decades running up to the genocide, one of the key mechanisms through which the conditions of structural violence (marginalisation, alienation, humiliation) that preceded and underpinned the genocide were disseminated and consolidated. Decentralisation, like participation, is not, on its own, necessarily a good thing. It depends on the underlying aims, ambitions and motivations of its adherents and supporters. While it can certainly oppress, subjugate and alienate communities, leading to frustration, anger and physical revolt, it can also engage communities by opening up new political spaces and renewing the social contract between citizens and their leaders while building social cohesion and stability. The important thing is to learn from and not replicate history.