We conducted two matching grant experiments with an international development charity. The first and primary experiment tests a matching grant from the Bill and Melinda Gates Foundation (BMGF) compared to a matching grant from an anonymous donor. The second, auxiliary experiment, establishes that the matching grant from BMGF in this context does generate further donations compared to a control. We find that naming BMGF as the matching donor raises more money, both compared to an anonymous donor and compared to control. In a key result, we find that the effect persists after the matching period, and that the naming-BMGF effect is heterogeneous—largest for donors who previously gave to other poverty-oriented charities. Combining this with a survey of representative Americans that shows a correlation between giving to poverty charities and familiarity with the BMGF, we conclude that the matching gift here primarily works through a quality signal mechanism.