The paper explores whether the presence of refugees increases the likelihood of civil conflict. The paper includes a good summary of the literature on the relationship between refugees and civil conflict, including the mechanisms for how refugees might spread conflict. Using new geo-coded data on refugee sites (formal camps and informal settlements) and civil conflict data at the subnational level from 1989 to 2008, the authors find no evidence that the presence of refugee sites generates or sustains civil conflict either in the province where the refugee site is located or in other provinces within the same country. Where refugees are concentrated within a single province of a country, which is the case in most refugee-hosting countries in any given year over the study period, refugee-hosting provinces actually experience large decreases in their likelihood of civil conflict. The authors suggest that this “conditional risk reduction effect” is probably due to increased state presence and humanitarian activities (i.e. increased infrastructure, assistance, and security) that are focused in refugee-hosting provinces because refugees are not dispersed throughout the country. Where refugees are dispersed throughout the host country, there is still no effect on civil conflict. These findings contradict conventional wisdom and existing quantitative research, which find that increased refugee populations from neighboring countries are associated with increased likelihood of conflict onset.