Under what conditions can UN peacekeeping deployments help resolve armed conflicts and minimize human suffering? Existing studies on this question have struggled with the fact that peacekeeping data tends to be highly aggregated, which makes it difficult to assess whether peacekeeping troops on the ground in fact serve as security guarantors. Against this background we present spatially and temporally highly disaggregated data from South Sudan, including information about troop strength on a daily basis and at the level of local UN peacekeeping deployment sites. The data was constructed from original UN command documents not otherwise available and allows for analysis at a more granular level than any other available dataset on UN mission strength. We show that the UN Mission in South Sudan
(UNMISS) sharply contracted in terms of its spatial presence immediately after the civil war broke out, despite the fact that the UN Security Council more than doubled the mission's mandated size at that time. We further show that there appears to be no empirical association between local deployment patterns and fatal events, whether assessed in terms of the possible effect of troop deployments on later fatalities or, vice versa, in terms of the responsiveness of mission deployments to earlier deadly violence. Strikingly, UNMISS appears largely disengaged from the violence that has ravaged South Sudan.