Although researchers have recently made some progress in explaining the outputs of the US economic aid decision-making process, their efforts to explain the allocation of US military aid have been rather disappointing. In this article, we follow previous studies that have assumed a two-stage process leading up to the allocation of military aid, while making three significant improvements over those efforts. First we employ a better model, including a variety of political, strategic, economic, and humanitarian variables we hypothesize to be related to the allocation of military aid. Second, we solve the `low n' difficulty that plagued previous research, by employing an extensive dataset that covers a global sample of countries covering the 1983-8 period. And third, unlike previous research where a two-stage process has been assumed, we employ a methodology that solves the difficulties associated with selection bias, which arises when two interrelated decisions are modeled separately. As a result of these improvements our results are much stronger than those of previous studies. We find that strategic, political, and economic interests, as well as human rights concerns and economic development, have been considered in the US military aid decision-making process.