Partial peace agreements with a subset of non-state actors can disrupt existing power dynamics within a conflict, potentially triggering new episodes of violence. Using a regression discontinuity approach, the analysis reveals a statistically significant increase in violence perpetrated by non-state armed groups, other than the FARC, in municipalities that rejected the peace agreement between the Colombian government and the FARC insurgency in the 2016 peace referendum. I interpret the peace agreement as an economic shock that altered the incentives for violence among non-state armed groups that did not participate in the peace process. Consistent with this interpretation, I find that the increase in violence is more pronounced in areas characterized by significant coca cultivation and gold mining activities.