The article seeks to resolve the debate in the empirical literature over the effect of arms races on dispute escalation. Until now, tests of this link have remained inconclusive due to the conduct of largely parallel, rather than integrative, research. Michael Wallace's 1979 study finding a strong link between arming and escalation suffered substantial criticism on grounds that both his index and his dispute sample were flawed. Utilizing the arms races measures offered by Diehl and Horn, this article delves into the conceptual, measurement, and sampling controversies that have made resolution before this impossible. The results of resetting the proposition indicate that regardless of the arms race measure or dispute set used, the relationship between arming and the escalation of disputes to war is positive and significant. Removal of controversial dyads from the World Wars does not alter the findings substantially (or in the assumed direction). By introducing a time lag, it is also shown that many of the disputes that both Wallace and Diehl designated as non-escalating arms races were between countries that were at war with each other within five years. Consequently, the conclusion is that arms races are strongly associated with the escalation of disputes to war.