The evaluation of climate policy has emerged as an important application of cost-benefit analysis (CBA). But the tool, as it had been most widely used previously, was not suited for this problem. Spatially, the effects of climate change transcend national boundaries and, temporally, they transcend generational time scales. CBA, in its standard form, relies on assumptions that are not fully appropriate in this context. In this essay, I discuss the shortcomings of CBA framed by its historical development and argue that its relatively recent application to climate change has contributed to growth in the literature re-evaluating its normative foundations. In relation to discounting, recent innovations emphasise aspects such as the aggregation of a multiplicity of time preferences and the distinction between inter-and intra-generational time preferences, which lead to discounting approaches that differ from the standard approach with the Ramsey equation. In relation to spatial equity, much of the recent literature argues for the use of equity weights to account for impacts on different income groups. These are significant changes relative to the orthodoxy in the field before the emergence of climate change.