Although a carbon value has often been integrated in the frameworks established to guide public decision-making, benefit-cost analysis (BCA) has played no more than a minor role in the design of climate policies. It is urgently necessary to promote BCA in this area, and there is currently a unique opportunity for doing so. Major countries are designing new packages in order to meet their commitments, as illustrated by the European Green Deal, recent decisions on the part of the Biden Administration, and the creation of a Chinese national carbon market. These constructive processes must be based on BCA. BCA is absolutely necessary in order to achieve net-zero emissions by 2050 at a reasonable cost. Indeed, abatement costs across and within sectors, and across and within countries, are extremely heterogeneous, and many of the policy instruments in use (subsidies, feed-in tariffs, technical standards, etc.) overlap inefficiently. The instrumental debate between carbon pricing and other instruments is sterile if it merely remains at the level of stating principles. BCA can help on this point too, by specifying comparisons between alternatives, identifying complementarities, and selecting the most relevant combinations of instruments. Its scope should therefore range from setting benchmarks for carbon pricing to assessing, e.g., green investments or measures to enhance carbon sinks. When applied to decarbonization policies, BCA requires firstly the selection of a carbon value, in order to monetize the climate benefits of investments and policies. However, the whole assessment framework must be updated, including the time horizon, the discount rate, the cobenefits of climate mitigation actions, and the pricing of climate risks. We show that such an updated framework leads to an upward revision in the assessment of the climate benefits of mitigation actions, and that combining the valuation of damages and cost-effectiveness approaches is necessary in order to meet the needs of policy assessment. Finally, there is a need to extend analysis beyond the efficiency criterion in order to deal with other dimensions of climate policies, particularly their distributive impacts. This requires specific analyses, which should be articulated with BCA and carried out at an early stage for a better implementation of climate policies than we have seen to date.