Ethnic conflict is the most common type of internal armed conflict in the world. It often involves systematic attacks on civilian populations and is therefore also the major source of most of the world's 9.2 million external refugees and 25 million internal refugees. In 2003, Asia-Pacific was the region second most affected by conflict-induced displacement and in 2004 it had the second largest global number of internal refugees following Africa. Given the likelihood that this trend will continue, it is perhaps surprising that a relative lack of research has been conducted concerning the relationship between ethnic conflict and refugee movements within this region compared to other areas. It is therefore imperative that a comprehensive study be undertaken to fill this void of knowledge. The fundamental question posed by my thesis is why do some ethnic conflicts produce external refugees and others do not in the Southeast Asia/Pacific region? To answer this question, this thesis develops a theoretical model from which to analyse variations in both external and internal refugee numbers as a result of ethnic conflict in the region. It applies the model to specific ethnic conflicts in Fiji, Indonesia, the Philippines, and the Solomon Islands during the period 1995 2005 and identifies a common set of factors conducive to creating internal and external refugees. The findings emphasise the interlinked nature of the variables and demonstrate that no single-factor explanation exists that can explain how refugees are created.