Canonical theories of the state argue that order promotes private investment: conflict creates fears that future returns will be destroyed or expropriated. We systematically review three decades of research quantifying the effect of armed conflict on investment activity. This work overwhelmingly focuses theoretically and empirically on the national-level impact of conflict on investment. Yet, investment levels reflect the decisions of individual firms, and companies operating within the same country can be differentially affected by conflict. We develop a theoretical framework to predict how firms respond to conflict based on how proximate their operations are to violence. Conflict may deter investment by disrupting production or raising policy uncertainty. By contrast, conflict may encourage investment where it hampers taxation and regulation. We then use firm-level panel data from the mining sector to estimate how firms’ with differential exposure to armed conflict change their investment levels. Our analysis reveals divergent responses: we find that firms with operations at conflict sites dramatically reduce their investments following violence. Yet, firms operating in the territory surrounding conflict, but at a remove from the fighting, actually increase their investment. Firms well-removed from violence see a small negative effect. Our analysis supports our theoretical mechanisms and enables us to overcome an ecological inference problem that bedevils prior cross-national research.