<p>Much of the literature on the relationship between military expenditure and economic growth overlooks the effects of military expenditure composition on the economy. This study examines the time-varying effects of U.S. military expenditure on U.S. economic growth based on disaggregated data using the maximum entropy bootstrap method, which not only accounts for variability but also yields more efficient and robust results from small samples compared to conventional inferences based on asymptotic distribution theory. Our analysis indicates that military expenditures generally impede economic growth. The results of a disaggregated analysis show that Research, Development, Test, and Evaluation spending has a positive effect on economic growth over time. Other military expenditure components generally have a negative effect. Furthermore, the impacts of military expenditure and its components on economic growth are heterogeneous across different periods.</p>