The last 55 million years has seen the largest perrisodactyls, rodents, and marsupials come and go. Some indricotheres weighed perhaps 20 tons. Alternatively, the horse and tapir families have seen evolutionary trajectories of increasing size. Ruminants begin small and hit their stride of size and diversity some 20 million years ago. Proboscideans have simply lumbered along. The seemingly regular turnover of large mammal groups and species, or changes of body size within a species seem associated with warming and cooling periods, and/or vegetation transitions. All seem tied to carbon dioxide levels. Here we combine three lines of research present in the literature to explain ebbs and flows of large body masses, and transitions between perrisodactyls and artiodactyls. First, high carbon dioxide levels may favor woodier vegetation of lower nutritional quality. Second, in terms of foraging allometries, herbivore body size may represent a tradeoff between finding (favors smaller size) and handling food (favors larger size). Third, foragers should have higher giving-up densities (GUDs) on lower quality foods, and when temperatures are lower. Together these can produce sometimes opposing eco-evolutionary feedbacks. The general trending downwards of carbon dioxide may explain the transition from giant perrisodactyls to artiodactyl communities. Furthermore, mid-epochal rises in carbon dioxide may select for new, larger species of browsers. Finally, the Pleistocene ice ages may emerge from feedbacks of C₃/C₄ plant communities, temperature, carbon dioxide and subsequent body size evolution. Foraging economics, digestion physiologies, and vegetation-climate feedbacks may drive the comings and goings of giant mammals.