I advance and test a theoretical argument of how participation in UN peacekeeping affects the likelihood of coup attempts in troop-contributing countries (TCCs). The argument highlights the interplay between the economic incentives of militaries in poor TCCs and the UN's preference for contributors with stable civil-military relations. Fearing the loss of UN reimbursement funds, militaries for which such funds are important will avoid visible acts of military insubordination, such as coup attempts, that would place their future participation in UN peacekeeping at risk. I test this proposition against time-series cross-sectional data on 157 countries in the 1991-2013 period using panel regression and matching. The data show that countries where the armed forces are more dependent on peacekeeping revenues experience fewer coup attempts than comparable peers, even when taking coup-proofing measures and other alternative explanations into account. I also find that the coup-restraining effect is only observed in periods where member states contribute enough troops to award the UN a real choice over alternative contributors. The study introduces a novel theoretic logic, presents empirical results at odds with the existing literature, and suggests important policy implications with regard to UN vetting and standards for troop-contributing countries.