What can states expect to receive in return for the military aid they provide to other states? Can military aid buy recipient state compliance with donor objectives? In this study, we systematically investigate the effects of US military assistance on recipient state behavior toward the United States. We build on existing literature by creating three explicit theoretical models, employing a new measure of cooperation generated from events data, and controlling for preference similarity, so that our results capture the influence military aid has on recipient state behavior independent of any dyadic predisposition toward cooperation or conflict. We test seven hypotheses using a combination of simultaneous equation, cross-sectional time series, and Heckman selection models. We find that, with limited exceptions, increasing levels of US military aid significantly reduce cooperative foreign policy behavior with the United States. US reaction to recipient state behavior is also counterintuitive; instead of using a carrot-and-stick approach to military aid allocations, our results show that recipient state cooperation is likely to lead to subsequent reductions in US military assistance.