This article presents a new data set on self-determination movements (SDMs) with universal coverage for the period from 1945 to 2012. The data set corrects the selection bias that characterizes previous efforts to code SDMs and significantly expands coverage relative to the extant literature. For a random sample of cases, we add information on state–movement interactions and several attributes of SDM groups. The data can be used to study the causes of SDMs, the escalation of self-determination (SD) conflicts over time, and several other theoretical arguments concerning separatist conflict that have previously been tested with incomplete or inferior data. We demonstrate the usefulness of the new data set by revisiting Barbara Walter’s influential argument that governments will not accommodate SD challengers if they face several potential future challengers down the road because they want to build a reputation for strength. We do not find support for Walter’s reputational theory of separatist conflict.