Research on power-sharing in Africa remains silent on the effects of national peace agreements on the sub-national level. Conversely, most armed conflicts originate and are fought in (or over) specific areas. A plausible hypothesis would be that for power-sharing to have the desired pacifying effect throughout the national territory, it needs to be extended to the local level. Based on fieldwork in six former hotspots in Liberia, Burundi and the Democratic Republic of Congo (DRC) we find that there is hardly any local content, including local power-sharing, in national agreements. However, contrary to our hypothesis, neither local content (inclusion of actors or interest) nor local-power-sharing (either introducing a local power balance or monopoly) are indispensable to effectively bring about local peace, at least in the short-term. On the contrary, it might even endanger the peace process. The importance of the sub-national level is overestimated in some cases and romanticised in others. However, the history of spatial-political links, centralised policies, and the establishment of local balances or monopolies of power ultimately play an important role.