Abstract The aid effectiveness literature (AEL) consists of empirical macroeconomic estimates of the effects of development aid. By the end of 2004, it comprised 97 econometric studies of three families of related effects. Each family has been analyzed in a separate meta-analysis. The AEL is an ideal subject for meta-analysis as it uses only a few formally similar models to estimate the same underlying effects. It is also an area with strong beliefs, often generated by altruism. When this whole literature is examined, a clear pattern emerges. After 40 years of development aid, the preponderance of the evidence indicates that aid has not been effective. We show that the distribution of results is significantly asymmetric reflecting the reluctance of the research community to publish negative results. The Dutch disease effect on exchange rates provides a plausible explanation for the observed aid ineffectiveness.