What explains how international military coalitions perform in war? Scholarship on coalition operations and conventional wisdom assert that "unity of command'" tends to lead to victory. Yet, battlefield military effectiveness to which unified command can give rise may be insufficient to generate the political conditions necessary to give rise to the level of sustained international cooperation that leads coalitions to cohere throughout a war. Alliance politics and international institutions literatures argue the distribution of bargaining power within inter-state collectives plays an important role in determining the quality of cooperation in peacetime. Scholars understand relatively less about how factors related to bargaining power, like shared interests, military capabilities, and institutions, affect the quality of coalitions' strategic decisions once war begins. To advance a novel account of variation in the strategic performance of inter-state coalitions in war, I develop a bargaining theory of strategy-making. The extent to which coalitions achieve wartime political objectives rests on the quality of the bargains they reach over the objectives, resources, and decision-making structures (i.e. command and control systems) that comprise military strategy. Asymmetrically distributed bargaining power inside coalitions gives rise to a wider range of bargaining problems than in symmetric coalitions. Multinational command systems function as international institutions, providing sources of credible information that help solve some of these bargaining problems while exacerbating others. Through different mechanisms explored in the dissertation, variation in command institutions' design features interact with the distribution of bargaining power, improving strategic performance in coalitions with symmetric distributions of bargaining power, but hampering performance in others. The dissertation develops the theoretical argument, presents a novel conceptual framework to systematically characterize variation in coalition command institutions, and introduces an original data set, the Wartime Inter-state Coalition Command Institutions Data Set, tracking such variation at the war front-side-day level across coalition wars from 1900 to 2003. Empirical tests of the theory include statistical analyses of front-level outcomes using the data-set, a process-tracing analysis of Anglo-American strategic decision-making during World War II based on original archival research, and a comparative case analysis of strategic outcomes across other Allied ground campaigns in WWII. The study illuminates unintended dynamics to which decisions over command system design can give rise and points the way to potential solutions.