Poverty comparisons - such as whether poverty has increased, or where it is greatest, are typically clouded in conceptual and methodological uncertainties. How should individual well-being be assessed in deciding who is poor? Is a household survey a reliable guide? Where should the poverty line be drawn, and does the choice matter? This monograph surveys the issues that need to be considered in answering these questions, providing an accessible introduction to the most recent literature. The strengths and weaknesses of past methods are discussed, and a summary of methodological recommendations is given. A number of new analytical tools are described which can greatly facilitate poverty comparisons, recognising the uncertainties involved.