Following the oil spills due to the accidents involving the oil tankers Erika and Prestige on the Atlantic coast, thermal remediation of oil polluted sands in a fluidized bed was studied. A thermal process is proposed as an alternative to the physico-chemical washing that has been carried out with the Erika wastes. The study consists of: choice of the polluted wastes; air fluidization behavior of oil polluted sands in a cold fluidized bed; depollution characterization in a hot fluidized bed; design and engineering of an industrial scale furnace and investment and operating cost evaluation of such a remediation unit. Polluted sands were prepared for different combinations of sand particle diameter, oil and water content. Fluidizability of oil polluted sands was checked in a cold model fluidized bed facility. Hot tests showed that combustion started as soon as the fluidized bed temperature was equal to or higher than 700 degrees C. The remediation of the polluted sand was good in all the cases with a remaining pollution in the sand of about 50 mg/kg. Flue gas emissions were within the regulatory limits. There was no danger of any ash fusion noticed even in the long duration tests. An industrial scale furnace was then designed based on the measured limits of fluidizability and the evaluated time necessary to get an almost complete remediation of the oil. Fluidized bed treatment costs were then evaluated and compared to the physico-chemical separation that has been used in the case of Erika oil spill and the thermal treatment in a rotating furnace. Fluidized bed combustion appeared to be one of the best options with a cost of only 80 EURO/t of polluted waste. (c) 2006 Elsevier B.V. All rights reserved.