The herbicide isoxaflutole 1 (5-cyclopropyl-4-isoxazolyl)[2-(methylsulfonyl)-4-(trifluoromethyl)phenyl]-methanone) has been applied preemergence at the rate of 125 g ha(-1) on corn crops grown on fields located in regions different as to their soil textures. Its metabolite diketonitrile 2 (2-cyano-3-cyclopropyl-1-(2-methylsulfonyl-4-trifluoromethylphenyl)propane-1,3-dione)-which is the herbicide's active compound-and its nonherbicide metabolite 3 (2-methylsulfonyl-4-trifluoromethylbenzoic acid) were measured in the 0-10 cm surface soil layer of the corn crops after the treatment and until the harvest. At the opposite of what occurred in plant shoots, the transformation of isoxaflutole 1 into diketonitrile 2 was not immediate in soil. In the 0-10 cm surface soil layer, this transformation occurred progressively according to an apparent second-order kinetics, and the soil half-lives of isoxaflutole 1 self were comprised between 9 and 18 days. The adsorption of isoxaflutole 1 onto the solid phase of the soil and its organic matter should explain the stabilization effect of soil, increased by the application of fresh organic fertilizer. The sum of the concentrations of isoxaflutole 1 and diketonitrile 2 disappeared in the 0-10 cm surface soil layer according to an apparent first-order kinetics, and the soil half-lives of this sum were comprised between 45 and 65 days. The sum of the concentrations of isoxaflutole 1 and of its metabolites diketonitrile 2 and acid 3 did not account for the amount of isoxaflutole 1 applied. The discrepancy increased with the delay after the application, showing that the acid 3 was further metabolized in soil into common nontoxic products, and ultimately into CO2. The conjunction of the adsorption of isoxaflutole and its metabolites (which reduced their mobilities) onto the soil and its organic matter, and their further metabolism should explain why isoxaflutole and its metabolites were not detected in the 10-15 and 15-20 cm surface soil layers during the crops.