Metal mobility and availability in solid samples, such as soils, sediments and sludges, is frequently studied with sequential extraction procedures, which allow to partition the total metal content into classes of different reactivity. From the experimental point of view, each extraction step requires many practical operations, which can represent sources of error if they are not properly executed. The first part of the paper gives indication on how to perform sequential extractions, from sample pretreatment to extract analysis, in order to reduce experimental errors. The second part describes the main applications of sequential extraction schemes for the characterization and management of contaminated soils: the assessment of the risks associated to metal pollution, the evaluation of the feasibility and efficiency of remediation techniques, the study of soil properties and of metal behaviour and plant availability.