A central goal of evolutionary ecology is to elucidate processes responsible for community structure. In this vein, multiple organismal attributes have been shown to constrain food web structure (e.g. body size, niche breadth, and the type of functional response). While tissue stoichiometry (carbon to nutrient ratio) may also influence food web structure, less is known of its role in food webs. Furthermore, a focus on organismal stoichiometry is critical in the context of changing ecosystem stoichiometry as a result of anthropogenic carbon emissions. We address this link between ecosystem stoichiometry, food web structure, and tissue stoichiometry with a process-based adaptive dynamics model. In our model, predator and prey stoichiometry coevolve in the context of ecological interactions. We find that: 1) Diversity increases asymptotically with carbon content at high nutrient availability 2) Directional selection reduces tissue stoichiometry, 3) Tissue stoichiometry is released from selection at high ecosystem stoichiometry with low nutrients, and 4) Diversity increases following carbon addition to food webs at equilibrium. These results suggest that phenotypic diversity can be determined in part by ecosystem stoichiometry, but that this relationship is not straightforward. For example, the number of phenotypes increased with carbon, but only at high nutrient availability. Atmospheric carbon accumulation may thus lead to diversification in nutrient rich ecosystems. In contrast, phenotypic variation should be greatest in ecosystems with high carbon and low nutrient content, as these ecosystems impose little selection on tissue stoichiometry. We have shown that ecosystem stoichiometry strongly influences the coevolution of organismal stoichiometry. Hence, a process-based understanding of food web structure must explicitly consider ecosystem and organismal stoichiometry.