Molecular hydrogen production occurs through the serpentinization of mantle peridotite exhumed at mid-ocean ridges. Hydrogen is considered essential to sustain microbial life in the subsurface; however, estimates of hydrogen flux through geological time are unknown. Here we present a model of the primary, abiotic production of molecular hydrogen from the serpentinization of oceanic lithosphere using full-plate tectonic reconstructions for the last 200 Ma. We find significant variability in hydrogen fluxes (1-70 • 10 16 mol/Ma or 0.2-14.1 • 10 5 Mt/a), which are a function of the sensitivity of evolving ocean basins to spreading rates and can be correlated with the opening of key ocean basins during the breakup of Pangea. We suggest that the primary driver of this hydrogen flux is the continental reconfiguration during Pangea breakup, as this produces ocean basins more conducive to exhuming and exposing mantle peridotite at slow and ultraslow spreading ridges. Consequently, present-day flux estimates are~7 • 10 17 mol/Ma (1.4 • 10 6 Mt/a), driven primarily by the slow and ultraslow spreading ridges in the Atlantic, Indian, and Arctic oceans. As methane has also been sampled alongside hydrogen at hydrothermal vents, we estimate the methane flux using methane-to-hydrogen ratios from present-day hydrothermal vent fluids. These ratios suggest that methane flux ranges between 10 and 100% of the total hydrogen flux, although as the release of methane from these systems is still poorly understood, we suggest a lower estimate, equivalent to around 7-12 • 10 16 mol/Ma (1.1-1.9 • 10 7 Mt/Ma) of methane.Plain Language Summary Hydrogen gas is produced when mantle rocks are exposed and react with ocean water at slow spreading mid-ocean ridges. Only these ridges tend to expose vast expanses of mantle rocks because the temperature is too cool to generate sufficient melt to produce basaltic oceanic crust. Hydrogen produced in this way is consumed by microbial life; however, there is no record of hydrogen through geological time. To overcome this we built a model approximating the volume of serpentinized mantle rocks produced at slow spreading ridges and used geochemical data to constrain the amount of hydrogen that can be produced, as the Fe (II)/[Fe (II) + Fe (III)] of serpentinite is a proxy of hydrogen production from serpentinization. In order to estimate the flux through time, we use global plate models which trace mid-ocean ridges and the spreading rate. We find that the volume of hydrogen produced is greatest when slow ridges are most abundant and that hydrogen production is concentrated from ridges in the Atlantic, Arctic, and Indian oceans. This is because when a supercontinent breaks up, it produces internal ocean basins that evolve slowly, relative to an ocean basin encompassing the supercontinent (i.e., the Pacific Ocean).