ABSTRACT Heavy metals can be essential micronutrients in trace concentrations but are often toxic at high concentrations. Physiologically stressful concentrations of heavy metals occur in natural, geological outcrops or result from human activities, such as mining and pollution. Although metal toxicity restricts the growth of sensitive species, some plants are more tolerant. The evolution of metal-tolerant populations has been demonstrated in several species growing on sites of anthropogenic metal contamination in the southeastern USA. The most common natural, metalliferous soils globally are serpentine soils derived from ultramafic rocks, which contain elevated nickel, cobalt, and chromium. The southeastern USA has few ultramafic exposures, but those at Soldiers Delight in Maryland and Buck Creek in North Carolina have been studied in some detail and are summarized here, including presentation of previously unpublished data. On both anthropogenic and natural sites, plants may tolerate edaphic metals by either exclusion or accumulation and sequestration. Plants known as hyperaccumulators take up and store metals at exceptionally high concentrations in their leaves. Hyperaccumulation is rare, with only about 580 species able to hyperaccumulate in the world. Only two examples of heavy-metal hyperaccumulation are known from the southeast: cobalt in Nyssa spp. and manganese in Phytolacca americana; however, these are unusual cases because the plants are widespread on nonmetalliferous soils and only occasionally occur on metal-contaminated soils where they can hyperaccumulate. In these situations, there is probably no selective advantage of hyperaccumulation, but it may occur as a side effect of other physiological processes.