Why do insurgent organizations sometimes kill civilians? Some do all the time, some do so occasionally, and some seem to never do so. Why? Recent research seeks to explain civilian victimization by insurgent groups, also called terrorism, focusing on factors such as insurgent weakness or country regime type. This book presents a different explanation, building on what is referred to as insurgent embeddedness--the extent to which an insurgent group is enmeshed in relationships with the state, other insurgents, and the public. With this framework, the book presents a set of propositions for why such a group might attack civilians. The book introduces new data on insurgent groups and their attributes (the Big, Allied, and Dangerous II Insurgency data) and empirically tests the arguments presented. Some factors robustly related to insurgent terrorism include government coercion, interorganizational alliances, interorganizational rivalry, and involvement in crime. In addition to discussing civilian targeting in general, we also examine how insurgent embeddedness might affect three important types of civilian victimization: terrorist attacks on schools, violence against the news media, and a proclivity for attacking the general public (not government officials or other symbolic types of targets). Finally, because the focus of this book is on relationships, an in-depth explanation of insurgent alliances and rivalries is presented, examining their changes and determinants over time. In total, the book provides a comprehensive look at how insurgent groups interact with other actors--and the implications for several types of bloodshed against civilians.