The main observation is that while RJ practices were developed in Belgium in a relatively generalized and coordinated fashion, the implementation of RJ has failed to elicit growth of these practices. Since the late 1980s RJ has been actively developed in both practice and policy, usually with the input of academic researchers. The first section reviews the RJ developments and practices in Belgium, highlighting the most important features, including several forms of victim-offender mediation, conferencing, and restorative programs in prisons. The author also highlights the role of legislation, the role of nongovernmental organizations (NGOs), and the role of academics as the driving forces behind the implementation of RJ practices in Belgium. The second section focuses more specifically on the institutionalization of RJ practices in Belgium and argues that institutionalization can be viewed from two perspectives: social potential on one hand and domination by the existing structures on the other hand. The third section analyzes the institutionalization of RJ practices in terms of crime policies and societal developments, such as the shift over the past decades from a welfare state to a participatory state and the emergence of the individual citizen. The author argues that the participatory principle of RJ is being applied in a restricted fashion because it is reduced to individual participation without impacting informal communities or public policies. In the final section, the tensions inherent in the institutionalization and positioning of RJ in Belgium are considered as the author contends that RJ should occupy an intermediate position that is semi-autonomous. This intermediate position would fall between the formal criminal justice on one hand and informal and community-oriented approaches on the other hand. In order to realize this position, RJ practices should be guided by an independent organizational structure. Note, references