Abstract: Federal regulators have often required environmental labels, which may be designed to help consumers to save money or to reduce externalities. Under prevailing executive orders, regulators are required to project the benefits and costs of such labels, and also to show that the benefits justify the costs. These projections can be extremely challenging, partly because of the difficulty of knowing how consumers will respond to labels, partly because of the challenging of converting behavioral changes into monetary equivalents. The benefits of environmental labels should include (1) the monetary value of the reduced externalities and (2) the monetary benefit to consumers, measured by willingness to pay. It may be difficult for regulators to know (1), and even if they can figure out (2), willingness to pay may not capture the welfare benefit to consumers, at least if consumers are not adequately informed (or if they suffer from some kind of behavioral bias). In principle, agencies should include, as part of (2), the moral convictions of people who care about environmental goods, at least if those convictions are backed by willingness to pay. In the face of the evident epistemic difficulties, sometimes the best that agencies can do is to engage in breakeven analysis, by which they explore what the benefits would have to be in order to justify the costs. Technical as they might seem, these claims raise fundamental questions about valuation of environmental goods and the possible disconnect between willingness to pay and welfare.