This paper addresses one of the empirical claims of the Ôresource curseÕ argument, namely that oil abundance raises the probability of political violence. I argue that the two main theoretical premises of the oil-civil-war link, rent-seeking and the rentier state model, fail to provide a convincing argument as to why oil economies are more vulnerable to the onset of civil war. I find that three often-mentioned mechanisms as to why oil economies succumb to political violence not convincing; that oil economies are poor economic performers; that oil economies generate high corruption; and that oil economies tend to produce authoritarian regimes. I also examine the empirical evidence on the oil-civil-war link and conclude that the results are not robust. I conclude with some policy implications.