Motivation: The Great Aid Debate is at an impasse between those who believe aid works and those who do not. The impasse is in part because people are arguing about the answers to different questions and in part because the methods historically used are not fit for purpose. The methods have not helped advance our understanding about how and why aid works. Research questions: Are there better questions and ways of measuring effectiveness that are better suited to developing and testing theories of aid, at a level more useful for policy-makers in promoting more effective development co-operation? Approach and methods: The article reviews the literature on aid effectiveness, development effectiveness and evaluation methods. It focuses on the questions these methods aimed to answer and the methods adopted. The literature broadly covers two questions at the macro level: "Does aid work?" and "How effectively is aid managed?" and a third question at the micro level: "How effective is each aid project?" Findings: A focus on the wrong questions and/or methods at the macro level has been at the expense of explicit aid theory as opposed to development theory. The micro level is too small to be useful in informing policy. A more useful question at a meso level is: "What types of development cooperation work in which contexts and why?" A trend of systematic reviews which rely on randomized control trials (RCTs) falls into similar traps as historical measures. Theory-based evaluation approaches are more explanatory as they investigate the mechanisms by which different types of development cooperation work. Policy implications: I propose that a meso level of analysis that sits above the level of individual initiatives and draws on the explanatory powers of theory-based evaluation and mechanisms will yield more useful theories about how different types of development cooperation work. It will also provide development cooperation policy-makers insights into what types of development cooperation are more likely to work effectively in different contexts, potentially allowing them to make better decisions.