The concept of 'safe areas' emerged in the early 1990s as a way of responding to increasing displacement triggered by internal conflicts. As a form of protection, their record was mixed - for every success like northern Iraq in 1991, there was a failure like the collapse of the Srebrenica safe area in 1995. But why did the safe area concept itself emerge at this time? Traditionally, safe areas were akin to humanitarian spaces anchored in consent. The shift in the early 1990s was to replace consent with an internati onal military presence, including military forces and peacekeepers. This article argues that this shift was only possible because of two critical changes which occurred within the United Nations: the recognition that civilian protection represented an international problem and the un Security Council broadening how it interpreted the notion of 'threats to international peace and security' to include issues such as forced migrant flows.