Understanding Earth’s tectonic processes is contingent on constraining their rates and durations. The timescale of metamorphism (i.e., for how long a rock is subjected to high pressure and temperature) is indicative of how heat and mass transfer during a tectonic event. During metamorphism, compositional gradients form in minerals and are modified by chemical diffusion. Forward modeling of the extent of diffusion quantifies the time involved in the production and preservation of these gradients, and thus proves to be a powerful tool for constraining metamorphic timescales (“diffusion chronometry”). However, tectonic settings of different types and ages have not been evenly targeted for rigorous diffusion studies. This thesis applies diffusion chronometry in garnet to various terranes and demonstrates its potential in addressing critical questions about Earth’s tectonics. The first chapter reviews the significance of timescale constraints in the study of tectonics and introduces the fundamental principles of chemical diffusion in garnet. Chapters 2 and 3 evaluate the secular change of global metamorphic cooling rates over Earth’s history using two case studies of an Archean craton, the Superior Province of North America. The Minnesota River Valley Subprovince is characterized by two neighboring crustal blocks that were metamorphosed contemporaneously to different grades by an advective heating event. They record strikingly different cooling rates that suggest greater complexities in the cooling histories of Precambrian orogens than commonly assumed. A comprehensive study of the Quetico Subprovince that contrasts diffusion chronometry with radiometric dating (“thermochronology”) further demonstrates the uncertainty and variability of Archean metamorphic cooling rates. The suggestion that the apparent increase in cooling rates globally, since the Archean eon, reflects fundamental tectonic changes should be evaluated with caution, given the inherent limitations and biases of existing data. Chapter 4 presents the first application of diffusion chronometry to constrain the timescales of material cycling deep in a subduction zone using complexly zoned garnet crystals from Jurassic subduction mélanges of Cedros Island, Baja California, Mexico. The pressure--temperature--time evolution of the subducted blocks cannot be explained by large-scale distributed flow in the subduction channel, as proposed in some numerical models. Instead, the subducted materials experienced more complex circulation and rapid exhumation via focused return flow.