Models of cadmium uptake and accumulation by plants are potentially of value in evaluating both the soil and plant components, in order that limits can be established, particularly in relation to soil concentrations, to restrict cadmium in the human food chain. Although the theoretical pattern of uptake is believed to follow a Mitserlich function, the best estimates to date that are based on empirical data use only linear functions, due to a paucity of suitable data, with r(2) values of 0.35 and 0.62 for maize and ryegrass, respectively. A feature of these empirical models is the strong dependence of cadmium uptake on the interaction between soil cadmium concentration and soil pH. Insufficient data currently exist to develop empirical models of foliar cadmium uptake. Mechanistic models are not well developed, but the most realistic models assume that ions are transported to roots by mass flow and diffusion and are absorbed at rates that depend on their concentration at the root surface, following Michaelis-Menten kinetics. However, the level of agreement with empirical data is unsatisfactory and further work is required to isolate the key variables contributing to the errors, in particular the role of ectomicorrhizal fungi on root uptake. The influences of root exudates and soil temperature on the solubility of different Cd species in the rhizosphere of apical root zones also require more detailed evaluation before incorporation into mechanistic models, and the absence of an accurate technique for estimating root surface area is an impediment. A further disadvantage of existing mechanistic models is the necessity for difficult and expensive root measurements, restricting their value for field predictions. Mechanistic foliar cadmium uptake models have been developed, but key variables such as translocation of heavy metals into the plant and the resuspension of the pollutants into the atmosphere have so far been ignored. Differential adherence of wet and dry particles to the leaves and the influence of soil splash on stem and leaf uptake remain to be effectively quantified. Although the parameters described so far may be generalized for a number of different soil and plant systems, the differences between genotypes on phytochelatin production and differential translocation rates cause major variation in accumulation rates and will ensure that both empirical and mechanistic models are genotype specific. Cadmium transport within the plant can be effectively dichotomized into short distance transport from the root to the stele, assumed to be a symplastic transport across the root cortex, and long distance transport to the shoots, mainly in the ionic form in xylem and phloem. Uptake into seeds is not well understood, even though it is of major importance for uptake into cereal grains. It is concluded that there is currently only a limited understanding and quantification of key parameters which would allow a comprehensive mechanistic model of C uptake by different plant genotypes to be constructed, and also that there is a limited number of empirical observations of key endpoints for an empirical model. Further work on these aspects is essential to facilitate the construction of effective models to control excessive Cd accumulation in the human food chain. (C) 2004 by Elsevier Inc.