Uptake of 54Mn by excised oat roots from dilute manganese chloride solutions has been investigated. The time-course of uptake has been analysed into the customary but somewhat arbitrary fast and slow phases. Uptake is not metabolic in either of these. The fast phase ('exchangeable' manganese) is essentially complete in about 30 minutes and represents the attainment of equilibrium in a process of ion-exchange. It is shown that analyses appropriate for enzyme kinetics cannot be applied in this situation, and an alternative formulation is based on Donnan equilibration, taking account of the selectivity of the ion exchanger towards different counter-ions; the predictions of this latter theory are compared with the experimentally determined uptake. The slow phase ('absorbed' manganese) may also involve exchange sites, either chemically different from, or more difficult of access than, those involved in the fast phase, or both. Equilibrium was certainly not reached in three hours in this slow-phase process.Release of manganese, taken up by the roots from manganese chloride solutions, into calcium chloride solutions does not seem to be simply the reversal of uptake, particularly with very dilute solutions. This is particularly shown by the kinetics of uptake and release, uptake being a much faster process than release. Manganese may transfer from the first phase to the second phase, but there is no evidence that uptake by the roots proceeds in series from first to second phase. It is considered more likely that the two phases function independently, linked by the surrounding solution.