This paper offers the first systematic study of civilian victimization in the fight between the Turkish state and the Kurdish insurgency (the PKK), one of the longest lasting ethnic conflicts in the post WWII period. After a period of negotiations resulting in decline in violence with 2013, the conflict was rekindled with greater intensity in the summer of 2015. What are the temporal and spatial dynamics of civilian victimization in this conflict? Which parties target civilians more frequently? After answering these questions, the paper adopts a more theoretical approach and addresses the patterns explaining civilian victimization. How do political events such as electoral competition affect the motives of warring parties engage in more indiscriminate violence? How does the Syrian civil war characterized by very high levels of civilian deaths affect the pattern of conflict in Turkey? The paper is based on a unique dataset about violent events related to the Kurdish insurgency since January 1, 2000 until the end of 2016. The Kurdish Insurgency Violent Events (KIVE) dataset covers all events with fatalities, identify the different types of fatalities (security forces, insurgents, and civilians), records the exact day, location (province-district-neighborhood/village), number and type of fatalities, the type (i.e., firefight, bomb attack, etc.) and nature of (i.e., initiated by insurgents) violence. The dataset utilizes information from a plurality of sources with different biases, including the Turkish state and insurgents, Turkish media, pro-PKK media, and human rights associations in both Turkish and Kurdish. While this strategy of triangulation does not eliminate political biases in reporting and the fog of war contributes to significant levels of uncertainty, the KIVE offers very detailed and precise information making it possible to identify the long-term evolution of the armed conflict in Turkey. The dataset shows a significant increase in civilian casualties in 2015 and 2016 with the coming to violence to urban cities. More than 45 percent of all civilians who were killed between 2000 and 2016 lost their lives in these two years. The initial analysis also suggests that the targeting of civilians is closely linked to political balance of power as measured by the levels of electoral competition. Politically more competitive areas (i.e., districts) are less likely to witness high levels of civilian victimization. Furthermore, the Syrian civil war had spillover effects over the conflict in Turkey. The Kurdish question in Turkey is now very much intermingled with the Kurdish question in Syria.