In most of the recent literature on state building, the signi cance of war relies on the assumption that interstate disputes and civil wars are two independent types of con icts, with opposing effects on the development of state capacity; while external con icts imply unity and consensus, internal wars are divisive and destructive. This paper contributes to this literature by developing an alternative explanation based on the idea that the actors in these two types of con icts might be linked. A rst result shows that the strength of this link determines whether con icts fought against internal adversaries (i.e. civil wars) or external enemies (i.e. interstate disputes) complement or substitute each other, leading to larger or smaller investments in scal capacity. The theory also predicts that political stability plays a non-trivial role in channelling the relation between both kinds of con icts and scal capacity. A second result shows that while less political stability translates into less scal capacity, more stability does not automatically imply more state-building. The exact relation depends on how cohesive institutions are. A novel mechanism through which more political stability might imply less state-building is proposed. Finally, I show that correlations in cross-country data are consistent with the theory.