Until the terrorist attacks of 11 September 2001, few aside from professional aid workers paid attention to one of the first major humanitarian crises of the twenty-first century. After over twenty years of continuous warfare, the Afghani population was nearly at the breaking point; more than 2 million Afghani refugees lived in Iran and Pakistan after fleeing civil battles and three years of drought. (1) Millions more were internally displaced inside Afghanistan, victims of ongoing struggles between the Northern Alliance and the Taliban, the Afghani ruling regime. After the terrorist attacks, world attention focused on the Afghani humanitarian crisis because the Taliban had a symbiotic relationship with Osama bin Laden and his organization al-Qaida, the alleged perpetrators of the attacks. As the United States and its coalition partners prepared for operations against Osama bin Laden, al-Qaida, and the Taliban, the potential for a humanitarian disaster of unprecedented dimensions quickly came to the atten tion of the world's press. By October 2001, with U.S. forces beginning military operations, a new wave of Afghani refugees began streaming into neighboring Pakistan and Iran. Although the United States pledged humanitarian assistance as part of its overall strategy, the international relief community was abandoning its operations on the ground. (2) Under the best of circumstances, aid workers and refugees in Afghanistan were insecure. Complicating relief efforts, the Taliban itself, alQaida loyalists, and Afghanistan's neighboring governments posed security risks to aid workers and refugees alike. Even before the current crisis, nongovernmental organization (NGO) relations with the nominal ruler of Afghanistan, the Taliban, were complicated by the religious affiliations of some NGOs. After the initial U.S. attacks, news reports indicated that Taliban members assaulted local UN aid workers. Worse still, aid workers were endangered by the U.S. air strikes themselves. In the fall of 2001, the situation in Afghanistan and neighboring countries exhibited all the signs of a potential complex humanitarian emergency by combining serious civilian suffering (from hunger, poor sanitation, and lack of shelter in the face of rapidly worsening weather) with concerns about the physical safety of both the refugees and aid workers. Yet, as with the Great Lakes region, Chechnya, and the various Balkan crises of the 1990s, the world community remained largely unprepared for the consequences of humanitarian disaster taking place in regions where political, religious, and ethnic conflicts might undermine the best efforts of humanitarian organizations. (3) Although international military intervention has often been justified on humanitarian grounds and many humanitarian NGOs receive government support, few formal measures have been undertaken by donor governments to increase the security of those providing relief to the victims of conflict. NGO aid workers are left on their own even when they operate, in effect, as proxies for international organizations and national governments. The same is also true when the military operations themselves lead to humanitarian emergencies, as in post-Gulf War Iraq. Traditionally, most humanitarian organizations have supported this hands-off approach because they have seldom believed that using military forces to protect aid workers is appropriate. In fact, many members of the humanitarian community believe that official involvement may make their jobs more difficult, exacerbate complex emergencies, and even endanger the local victims. Thus, despite the increased internationalization and privatization of humanitarian assist ance, the governments who fund relief activities have taken few measures to increase the security of humanitarian aid workers. In this article, we argue that this position is untenable in a world of failed states, gray area phenomena, increasingly violent internal conflicts, and, as recent events demonstrate, a global war on terror. …