AbstractHydrothermal convection at mid‐ocean ridges links the ocean's long‐term chemical evolution to solid earth processes, forms hydrothermal ore deposits, and sustains the unique chemosynthetic vent fauna. Yet the depth extent of hydrothermal cooling and the inseparably connected question of how the lower crust accretes remain poorly constrained. Here based on coupled models of crustal accretion and hydrothermal circulation, we provide new insights into which modes of lower crust formation and hydrothermal cooling are thermally viable and most consistent with observations at fast‐spreading ridges. We integrate numerical models with observations of melt lens depth, thermal structure, and melt fraction. Models matching all these observations always require a deep crustal‐scale hydrothermal flow component and less than 50% of the lower crust crystallizing in situ.