The identification of the occurence and results of warfare in the archaeological records of prehistoric societies has always been fraught with difficulties. This study investigates this prickly area through the development of a series of correlates, resulting in testable models which can be applied to the archaeological record. Correlates include causal correlates, such as climatic deterioration, or demographic rise; preparatory correlates, such as appearance of fortifications or proliferation of weaponry; functional correlates, such as burned sitesor unburied bodies/weapons trauma; and relative correlates, such as change in subsistence or trading practice. It is based on three case studies - the Middle Thames Region in the Later Bronze Age, Gallica Belgica, and the Salt River Area, Arizona, c.1250-c.1450 AD.