This presentation deals with the issue of modularization of chemical plants, defined in the sense of designing smaller-capacity facilities (“modules”) that could be used to process geographically-distributed resources or serve distributed customers. Using established scaling relations, it is demonstrated that technologies that scale up well (i.e., for which capital cost increases sublinearly with capacity) scale down poorly, and vice versa. A further analysis of scale-down and modularization is carried out from the perspective of scaling of individual unit operations and their contribution to the plant cost stack. Following the same line of reasoning, it is shown that units that scale down well will scale up poorly and vice-versa. On this basis, it is argued that scaling relations can be used not only to guide the design of modular processes based on existing technology, but also to direct research towards new designs that alleviate the cost problems associated with units that have poor scale-down characteristics. These concepts are briefly illustrated by analyzing the scale-down of the reactor present in an ammonia plant.