It is often assumed that policymakers and populations grow weary from war. The costs associated with the use of military force are thought to become readily apparent in the wake of devastating conflicts, and support for military forays is expected to erode as a result. Although the notion is compelling, to date, empirical studies of war-weariness have yielded mixed results. This article re-examines war-weariness theory. It focuses upon a recent, but largely neglected, refinement in the theory as well as insights gleaned from the foreign-policy learning literature. The theory is recast with emphasis placed on the collective weight of consecutive war wins and consecutive war losses. War- weariness is postulated to follow a series of war defeats. A series of victories is assumed to produce the opposite effect. Bolstered by past success, decisionmakers in states that have a run of war wins should use military force more frequently. To test the article's hypotheses, time series, cross-section negative binomial and OLS regressions of 66 states for the period 1946-96 are employed. Interstate military intervention is the type of force analyzed. Intervention initiations are found to not only increase as states win successive wars, they also increase as war losses mount. This U-shaped relationship holds for large-scale interventions as well. Although these results provide little empirical support for a cumulative approach to war-weariness, they do suggest that a state's war record affects decisions to intervene militarily. In the realm of low-scale military force, the weight of the past does affect policymakers and is important for understanding contemporary policy.