A crude contaminated soil, arising from an oil production zone in Tabasco, Mexico was studied. A sample of about 40 kg was dried and screened through meshes 10-100. Total petroleum hydrocarbons and 6 metals (Cd, Cu, Cr, Ni, V and Zn) were determined to the different portions. For soil which passed mesh 10, six non-ionic, three anionic and one zwitterionic surfactant solutions (0.5%) were employed to wash the soil. Additional tests using surfactant salt mixtures and surfactants mixtures were carried out. Once the best soil washing conditions were identified, these experimental conditions were applied for washing the rest of the soil portions obtained (meshes 4, 6, 20, 40, 60, 80, 100). Total petroleum hydrocarbons values were in the range of 51,550 to 192,130 mg/kg. Cd was not found in any of the soils portions, and the rest of the metals were found at different concentrations, for every soil mesh. Treatability tests applied to the soils indicated that it is possible to get removals between 9.1 to 20.5%. For the case of a sodium dodecyl sulphate 1% solution, total petroleum hydrocarbons removal was as high as 35.4%. Combinations of sodium docecyl sulphate and salts, gave removal rates up to 49.5%. Total petroleum hydrocarbons concentrations for the whole soil were about 150,600 mg/kg. The higher the particle size, the lower the washing removal rate. The combined effect of particle size and total petroleum hydrocarbons concentration, determines the total petroleum hydrocarbons removal efficiencies. These facts are very important for designing an appropriate soil washing remediation process.