Why did the US prefer multilateral alliances in Europe, but bilateral alliances in Asia after World War II? Rationalists and constructivists debate the impact of power, institutions, and identities in explaining this highly contested question. We introduce a new argument embedded in prospect theory from political psychology — a prospect–threat alliance model — to account for the variation in US alliance strategy toward Europe and Asia after World War II. Through setting the threat level as a reference point for leaders’ prospects of gains or losses, we suggest: (1) high threats frame decision-makers in a domain of losses, and multilateral alliances become a favorable alliance choice because states are more likely to take the risk of constraining their freedom of action in return for more help from multiple allies as well as for avoiding further strategic losses; (2) low threats position leaders in a domain of gains, and bilateral alliances win out because states are risk-averse in terms of maintaining their freedom of action in seeking security through alliances with fewer allies. US alliance policy toward Asia after World War II is a within-case analysis that tests the validity of the prospect-threat alliance model.