PREMISE OF THE STUDY: A recent commentary by Edwards et al. (Am. J. Bot. 103: 975–978) proposed that constraints imposed by the packing of young leaves in buds could explain the positive association between non-entire leaf margins and latitude but did not thoroughly consider alternative explanations. METHODS: We review the logic and evidence underlying six major hypotheses for the functional significance of marginal teeth, involving putative effects on (1) leaf cooling, (2) optimal support and supply of the areas served by major veins, (3) enhanced leaf-margin photosynthesis, (4) hydathodal function, (5) defense against herbivores, and (6) bud packing. KEY RESULTS: Theoretical and empirical problems undermine all hypotheses except the support–supply hypothesis, which implies that thinner leaves should have non-entire margins. Phylogenetically structured analyses across angiosperms, the El Yunque flora, and the genus Viburnum all demonstrate that non-entire margins are indeed more common in thinner leaves. Across angiosperms, the association of leaf thickness with non-entire leaf margins is stronger than that of latitude. CONCLUSION: We outline a synthetic model showing how biomechanics, hydraulics, vein geometry, rates of leaf expansion, and length of development within resting buds, all tied to leaf thickness, drive patterns in the distribution of entire vs. non-entire leaf margins. Our model accounts for dominance of entire margins in the tropics, Mediterranean scrub, and tundra, non-entire margins in cold temperate deciduous forests and tropical vines and early-successional trees, and entire leaf margins in monocots. Spinose-toothed leaves should be favored in short-statured evergreen trees and shrubs, primarily in Mediterranean scrub and related semiarid habitats.