We review methodological issues that arise in designing, implementing and evaluating the efficacy of 'light' paternalistic policies. In contrast to traditional 'heavy-handed' approaches to paternalism, light paternalistic policies aim to enhance individual choice without restricting it. Although light paternalism is a 'growth industry' in economics, a number of methodological issues that it raises have not been adequately addressed. The first issue is how a particular pattern of behavior should be judged as a mistake, and, relatedly, how the success of paternalistic policies designed to rectify such mistakes should be evaluated – i.e.,the welfare criterion that should be used to judge light paternalistic policies. Second,paternalism, and especially light paternalism, introduces new motives for attempting to understand the psychological processes underlying economic behavior. An enhanced understanding of process can help to explain why people make mistakes in the first place,and, more importantly, provide insights into what types of policies are likely to be effective in correcting the mistakes. Third, there is an acute need for testing different possible policies before implementing them on a large scale, which we argue is best done in the field rather than the lab. Fourth, in addition to methodological issues, there are pragmatic issues concerning who will implement light paternalistic policies, especially when they involve positive expenditures. We discuss how economic interests can be rechanneled to supportendeavors consistent with light paternalism.