The thesis explores the production of violence in the intervention in Libya in 2011. In March that year an uprising against then leader Muammar Gaddafi resulted in an international intervention authorised by the United Nations (UN), which involved a no-fly zone, and aerial bombing campaign to protect civilians in Libya. The intervention is seen as the first military intervention authorised by the UN to refer to the 'responsibility to protect' (R2P) norm. This thesis analyses the processes involved in the production of the international intervention in Libya through the categories of time, space and subjectivity. In doing so it draws upon neo-Foucauldian accounts of contemporary security practices, and postcolonial imaginative geographies of security and insecurity. My argument is threefold. The first argument concerns the political topography of intervention and R2P. It challenges claims of spatial separation in which political territories are clearly demarcated and separated, arguing that the security practices of liberal states are now predicated on a deterritorialised network of surveillance, monitoring and border control in which borders are not obsolete but are highly differential in impact. The second claim of the thesis concerns a displacement in the ethical measurement of humanitarian war to a temporal plane. The success of intervention is measured not in terms of lives saved, or the extent of rebuilding and reconstruction after conflict, or the humanitarian impact, but in the speed at which intervention is mobilised. The third central claim of the thesis is that humanitarian intervention depends upon, reproduces and perpetuates divisions and distinctions between people who are worthy of protection and those who may be killed. The moral universe of intervention is one where people are divided according to legitimacy based upon a series of assumptions about race, gender, class, religion and nationality. Those deemed legitimate potential subjects are worthy of saving, and many others are seen as illegitimate, and can be sacrificed without regret. The thesis concludes by arguing that the assumptions of humanitarian intervention serve to normalise violent responses to insecurity and crisis, reducing the space for non-violent alternatives. I argue that as a response, we should emphasise the possibilities for non-violent resistance.