The econometric literature has been unable to establish a robust association between foreign aid and growth and poverty reduction. In this paper we argue that aid effectiveness must be assessed using methods that go beyond cross-country regressions. We calibrate a dynamic general equilibrium model that is capable of generating large income gaps between rich and poor countries. The model quantifies three sources of (i) lack of access to international capital, (ii) low schooling and high fertility (a poverty trap), and (iii) antigrowth domestic fiscal policy. We analyze policies designed to address each source of poverty and estimate and compare the aid cost of implementing the different policies. The policies differ dramatically in the extent and timing of their growth effects, and in the aid cost of their implementation.