Alfalfa (Medicago sativa) plantlets were exposed to Cd or Hg to study the kinetics of diverse stress indexes. In the so-called beaker-size hydroponic system, plantlets were grown in 30 microM of Cd or Hg for 7 d. Oxidative stress took place and increased over time, a linear response being observed with Cd but not with Hg. To improve the sensitivity of the stress assays used, a micro-assay system, in which seedlings were exposed for 24 h, was developed. Phytotoxicity of metals, quantified as growth inhibition, was observed well before there was any change in the non-protein thiol tissue concentration. When measured with conventional techniques, oxidative stress indexes did not show significant variation. To trace early and small plant responses to Cd and Hg, a microscopic analysis with novel fluorescent dyes, which had not yet been exploited to any significant extent for use in plants, was conducted. These fluorescent probes, which allowed minute cellular responses to 0, 3, 10, and 30 microM of both metals to be visualized in the roots of the alfalfa seedlings, were: (i) 2',7'-dichlorofluorescin diacetate that labels peroxides; (ii) monochlorobimane that stains reduced glutathione/homoglutathione (GSH/hGSH); and (iii) propidium iodide that marks nuclei of dead cells. Oxidative stress and cell death increased after exposure for 6-24 h to Cd and Hg, but labelling of GSH/hGSH decreased acutely. This diminution might be the result of direct interaction of GSH/hGSH with both Cd and Hg, as inferred from an in vitro conjugation assay. Therefore, both Cd and Hg not only compromised severely the cellular redox homeostasis, but also caused cell necrosis. In plants treated with 1 mM L-buthionine sulphoximine, a potent inhibitor of GSH/hGSH synthesis, only the oxidative stress symptoms appeared, indicating that the depletion of the GSH/hGSH pool was not sufficient to promote cell death, and that other phytotoxic mechanisms might be involved.