Abstract: It is commonplace to think that composition is explained by relations—what I call the relational idea. Many argue that taking this seriously requires rejecting the idea that classical mereology—the most well-known formal theory of parthood—tells us anything important about the nature of composites. In this paper, I have two aims. First, I want to explore the extent to which we can take the relational idea seriously without abandoning classical mereology completely. Second, I want to motivate an interest in a neglected question about composition, which asks how relations figure in explanations of composition. I argue that by taking this question seriously, we can better understand the extent to which the relational idea is in tension with classical mereology.