A one-box model is used to examine the relationship between freshwater residence time ( T) and the time constant (half-life, t 1 2 ) for a first-order chemical reaction in an estuary. Higher and lower limits of the ratio t 1 2 /T appropriate for conservative or chemical equilibrium modelling, respectively, are derived. These limits depend also on the equilibrium constant for the reaction and the initial reactant and product concentrations. For intermediate values of this ratio, the system can be described quantitatively only by using a kinetic model. Systematic inter-estuarine variability in the distributions of dissolved metals as a function of estuarine size (or hydrodynamic turnover time) is predicted by this model when the potential for sorptive particle-water exchanges is examined. However, this deduction, which is based on kinetic criteria alone, is not confirmed by evidence in the literature. Indeed, the model is contradicted in that dissolved metals appear to be most interactive in some, but not all, small estuaries and least interactive in large estuaries. It follows that the thermodynamic drive for metal interactions to occur is a necessary additional consideration. Gross thermodynamic disequilibrium with respect to particle-water metal exchange, which in turn generates highly non-conservative dissolved metal distributions, is most readily generated in small, highly dynamic systems through enhanced internal mobility of estuarine resuspendable sediment.