The rate of dissolution of many rock-forming minerals is controlled by the hydrolysis of the bridging silicate bonds at the mineral surface. This hydrolysis can be studied at a detailed level by combining experiments in isotopically distinct solutions with molecular-orbital calculations of the reaction energetics and geometry. These calculations are linked to the macroscopic process of dissolution via the transitionstate theory. Rates of silicate leaching and dissolution in D 2O and H 2O differ for several reasons. First, hydrolysis involves transfer of hydrogen from the solution to a bridging siloxane bond at the mineral surface. The transition state equilibrium describing this reaction varies with the vibrational properties (and, hence, isotopic composition) of the reactant and the transition state. Secondly, the equilibrium acid-base properties of the oxide surface in H 2O and D 2O are not identical. These differences are important because rates of hydrolysis reactions are enhanced by adsorbed hydrogen and hydroxyl ions. On the basis of the molecular orbital calculations and an assumed mechanism of hydrolysis, quartz dissolution is predicted to be roughly a factor of four slower in D 2O than in H 2O at pH = pD = 3. The activation energy is predicted to be ≈20 kcal /mol, which is in agreement with values measured at hydrothermal conditions. At pH/pD conditions near the isoelectric point of quartz and in the temperature range 20-70°C, the measured rate in D 2O is only about 15% slower than in H 2O and the activation energy is ≈8 kcal/mol. The activation energy is slightly higher at pH/pD = 11 , but the kinetic isotope effect remains small. The discrepancy suggests that the hydrogen transfer to bridging oxygen from water at the experimental conditions is more rapid than modeled and may proceed early in the overall reaction.