After the end of the Cold War many post-socialist states had to face internal threats to their territorial integrity, since politically organized separatist groups within their boundaries started demanding for independence. Most ethnopolitical conflicts turned violent, whilst others were managed peacefully. What factors explain the different outcomes? The purpose of this comparative analysis is to explain how a high degree of access to executive power of separatist minorities can –if not avoid secession- at least prevent an armed conflict onset. I argue that any change in the group political status may influence the forms of mobilization. My main hypothesis is that if a group can submit its requests through institutional channels, it will be more prone to select conventional forms of mobilization and rely on party politics, rather than on extra-parliamentary activities. Moreover, in power sharing regimes actors can use their veto power in order to produce an institutional deadlock, making impossible the adoption of any federal policy. This situation forces the counterpart to seek a compromise with the separatist forces. Conversely, if a group systematically excluded from power its political leaders will try to mobilize the group against the state. Due to the absence of institutional channels to submit their demands, excluded groups have to select unconventional forms of collective action, thus making the resort to violence more likely. In order to explain how power-sharing influenced the dynamics of ethnopolitical conflicts, the whole population of secessionist conflicts -ended with the de jure or de facto secession of contested territories- occurred in post-socialist countries are included in this study. I will analyze both conflicts which turned violent (in South Caucasus, Moldova and Former Yugoslavia) and those who were managed peacefully (Montenegro and Slovakia).