This report draws on the new Amnesties, Conflict, and Peace Agreement (ACPA) dataset, developed through the Political Settlement Research Programme, to explore when and how amnesties are used during conflict and peace. In particular, it uses the empirical data from its large-scale comparative of analysis of amnesty practice to examine how the context in which amnesties are adopted can shape decisions on the scope, conditionality, and legal effects of amnesties. The report argues that these aspects of amnesty design can have significant implications for the extent to which amnesty can contribute to inclusive political settlements or conversely to excluding some individuals or groups from the post-conflict political contract.