Self-control and related concepts appear regularly in tax discussions, but often they are invoked hazily or blurred together with other aspects of choice over time. Despite the evident relevance of willpower to consumption patterns, wealth accumulation, and, ultimately, well-being, there is no consensus about whether and how heterogeneity along this dimension should factor into tax policy. There is support in the tax literature for such divergent responses as funneling more resources to low-willpower people, penalizing them for their lapses, and limiting their choices. Whether we should follow one of these approaches, or some other approach entirely, requires a careful analysis of willpower’s workings and its connections to well-being. To begin such an analysis, I focus on three categories of costs associated with willpower problems: the failure costs of suboptimal choices, exercise costs stemming from the willpower exertion itself, and erosion costs that relate to changes over time in willpower levels as a result of patterns of exertions and outcomes. With this framework in mind, I consider the effects of existing and proposed tax policy measures on people with different self-control levels. I then consider some alternatives that would address heterogeneity in willpower through a menu of regulatory bundles designed to induce self-sorting.