We investigate the impact of health aid on infant mortality conditional on the quality of governance in 96 recipient countries. Our analysis applies the long difference estimator and instrumental variable estimation, with aid instrumented by donor government fractionalization interacted with the probability of allocating health aid to a recipient nation. The effectiveness of health aid in reducing infant mortality is conditional on good governance (measured either as government effectiveness or control of corruption). Specifically, health aid to a recipient nation that experiences a one standard deviation improvement in government effectiveness reduces infant mortality by about 4 percent. Our findings reaffirm the importance of improving the quality of governance in recipient nations.