Two general families of geochemical models presently exist. The ion speciation-solubility group of geochemical models contain submodels to first calculate a distribution of aqueous species and to secondly test the hypothesis that the water is near equilibrium with particular solid phases. These models may or may not calculate the adsorption of dissolved constituents and simulate the dissolution and precipitation (mass transfer) of solid phases. Another family of geochemical models, the reaction path models, simulates the stepwise precipitation of solid phases as a result of reacting specified amounts of water and rock. Reaction path models first perform an aqueous speciation of the dissolved constituents of the water, test solubility hypotheses, then perform the reaction path modeling. Certain improvements in the present versions of these models would enhance their value and usefulness to applications in nuclear-waste isolation, etc. Mass-transfer calculations of limited extent are certainly within the capabilities of state-of-the-art models. However, the reaction path models require an expansion of their thermodynamic data bases and systematic validation before they are generally accepted.