This paper explores a 15-country household data base to evaluate the impact of three key assets (land, education and infrastructure) on rural poverty. Using both a descriptive analysis and a quadratic probit model, with the probability of being poor as a function of these three assets, the paper concludes that household access to education and infrastructure are positively associated with higher incomes, while the impact of land holdings varies across countries. Also, this paper shows the importance of the complementarities among assets in their poverty alleviating potential. The key policy implications revolve around long-term public investments in education and rural infrastructure, and asset-bundling, i.e. investing in assets simultaneously