In 2008, nearly half of Africa’s civil wars were ‘low-capability’, that is, conflicts without the tanks, artillery and jet-aircraft we associate with modern warfare. Although depictions of teenagers with AK-47s stealing aid supplies, looting diamonds and brutalising civilians are popular, we know little about the logic of combat in places such as Sierra Leone, Liberia and Somalia. This thesis advances our knowledge by articulating a theory of military strategy in low-capability civil wars. I argue that low-capability civil wars are defensive wars. Due to the inability of governments and insurgents to conduct offensive operations, the costs of capturing territory are far greater than holding it. Based on this theory, I predict that economic geography structures where battles are fought and when foreign states intervene in low-capability civil wars. Results from a quantitative analysis of African civil wars from 1960-2008 and a case study of the Liberian civil war show that the incidence and concentration of fighting around ‘point resources’, especially capital cities, is higher in low-capability civil wars when compared to the more familiar ‘conventional’ and ‘guerrilla’ civil wars. Foreign states also time their deployments with decisive battles over economically valuable areas in low-capability warfare, a pattern not observed in conventional or guerrilla warfare.