To date, little attention has been paid to the question of how episodes of mass killing are terminated. This has allowed several misconceptions, such as the notion that external armed intervention is a principal form of ending, to arise and profit. This study presents preliminary findings from a survey of cases of state perpetrated mass killing since 1945. It examines the forms of ending, finding that around half end only when the perpetrators themselves decide to end the killing, usually because they have accomplished their goals. It also explores the relationships between modes of termination and lethality and the resilience of different types of ending and offers insights into the implications for policy of some of these findings. It argues that foreign armed intervention is extremely rare and does not deserve, therefore, to be the common ‘go to’ option of advocates and analysts. Instead, presuming that armed intervention is off the table, it is more important to think in terms of what can be done to shape the perpetrators’ incentive structures or encourage internal dissent within the perpetrating elite. Finally, it shows that there is no easy overlap between what is morally palatable and what saves lives. Arming rebels may be morally pleasing, but may lead to protracted civil wars with atrocities – the worst of all outcomes. Likewise, negotiating to secure the state perpetrators’ core interests may feel immoral but might stop the killing and save lives.