Much information about the earliest stages of weathering comes from laboratory experiments where the mineral surface chemistry is rigorously controlled and the solution composition is maintained far from equilibrium with the solid. The experiments show that the pathways for dissolution are similar to those for ligand exchange around dissolved metal complexes. One difference is that the rates of mineral dissolution are controlled by the concentrations of surface species while rates of ligand exchange are proportional to bulk concentration. Nevertheless, the correlation is so strong that rates of olivine dissolution are predictable from the rates of solvent exchange around the corresponding divalent metal in solution. However, this level of resolution is achieved by maintaining unnaturally high fluid/mineral ratios and by taking great care to avoid precipitation of secondary minerals. Natural olivine weathers in intimate contact with secondary minerals. Reactions commonly proceed in clay-filled channels only a few tens of Ångströms in diameter and secondary minerals are topotactic with olivine. Primary and secondary minerals are so intimately intergrown in these channels that the olivine surface chemistry is undoubtably influenced by overlapping electrostatic double layers. Smectite growth may have proceeded at near-saturation with the fluid, but the solution is in gross disequilibrium with the olivine. There nevertheless remains some qualitative consistency between weathering in the laboratory and in the field. The ligand-exchange model for bond cleavage, for example, predicts that the presence of ferric iron at the mineral surface will retard weathering rates. Correspondingly, defect planes of oxidized olivine (to produce Fe 3+ and a vacancy) weather at a much slower rate than unoxidized material. Thus, the appropriate level of comparison of field and laboratory weathering is at the scale of reactivity trends; it is unreasonable to expect quantitative similarity in reaction rates.