There are two ways in which we might think about some particular distributive problem — for example, the fair distribution of costs associated with human caused climate change. We might think about it non-holistically and, in so doing, attempt to justify principles for the regulation of these costs considered in isolation. Or, we might think about it holistically by embedding the problem as part of a wider concern for distributive justice and, in so doing, attempt to generate principles which regulate all relevant benefits and burdens. In this presentation, I argue that it would be a mistake to adopt a non-holistic approach to problems of this sort.  Insofar as we want to think about any particular distributive problem, we should think about it in connection with all other distributive problems. Failing to do so is likely to generate morally undesirable principles. To make matters more concrete, I engage with a recent argument by Darrel Moellendorf, who employs a non-holistic approach within a Rawlsian framework to derive a principle for the just assignment of the intergenerational costs of climate change. I will demonstrate, however, that the problems associated with taking non-holistic approaches to distributive justice arise more generally.