Extra-Lethal Violence, a form of physical aggression that goes beyond the necessity to kill someone, presents a conundrum: it is inefficient and dangerous to produce, especially during warfare. Extra-Lethal Violence, particularly when it manifests in warfare, does not contribute to the immediate survival of individuals; the time, effort, and lack of awareness of surroundings or other attack suggests that Extra-Lethal Violence could be maladaptive at the individual level or in the short term. Yet this individually risky behavior that seems to have no direct benefit to the aggressor is both common and persistent across time and space. We utilized the electronic Human Relations Area Files (HRAF) to conduct a cross cultural analysis of the prevalence and potential predictors of Extra-Lethal Violence. Our research indicates that Extra-Lethal Violence is present across all populated regions of the world, for the entire timespan of the ethnographic record up to the ethnographic present, across subsistence, marriage, and social complexity levels. Our research suggests that Costly Signaling Theory (CST) is currently the best explanation for this behavior. Extra-Lethal Violence can be characterized as a difficult to fake, clear indicator of martial skill and physical fitness that has a high broadcast efficiency, both within and between groups. Rather than allowing behaviors such as Extra-Lethal Violence to be labeled as 'abhorrent' or 'disgusting,' we must view Extra-Lethal Violence in the same light as the cavalry, the ironclads, or nuclear weapons: societies seeking a decisive advantage over their enemies, utilizing available resources, be they material or behavioral.