The use of plants to clean-up soils contaminated with trace elements could provide a cheap and sustainable technology for bioremediation. Field trials suggested that the rate of contaminant removal using conventional plants and growth conditions is insufficient. The introduction of novel traits into high biomass plants in a transgenic approach is a promising strategy for the development of effective phytoremediation technologies. This has been exemplified by generating plants able to convert organic and ionic forms of mercury into the less toxic, volatile, elemental mercury, a trait that occurs naturally only in some bacteria and not at all in plants. The engineering of a phytoremediator plant requires the optimization of a number of processes, including trace element mobilization in the soil, uptake into the root, detoxification and allocation within the plant. A number of transgenic plants have been generated in an attempt to modify the tolerance, uptake or homeostasis of trace elements. The phenotypes of these plants provide important insights for the improvement of engineering strategies. A better understanding, both of micronutrient acquisition and homeostasis, and of the genetic, biochemical and physiological basis of metal hyperaccumulation in plants, will be of key importance for the success of phytoremediation.