Several simple models for the estimation of the half-life (t(1/2)) for the depletion of an organic chemical from a soil surface to air were examined. For moist surfaces, two models are proposed: the first requires knowledge of the soil/organic carbon partition coefficient (K(oc)) and the Henry's law constant (H) and the second the vapor pressure (P(s)) of the chemical involved. Due to uncertainties in the experimental K(oc) values those ones predicted by the group-contribution model of Meylan et al. [Environ. Sci. Technol. 26 (1992) 1560]-and proposed by the U.S. Environmental Protection Agency (EPA)-should be used. If reliable experimental P(s) values are not available, the first model is proposed, where in cases when H values are not available, predicted ones by the Bond-Contribution method of Meylan and Howard [Environ. Toxicol. Chem. 10 (1991) 1283]-and also proposed by EPA-can be used. In general, the agreement of the predicted t(1/2) values with the measured ones is within a factor of 3-5. Similar expressions, but with somewhat poorer results, are presented for dry field soils. In all cases, the obtained results represent a substantial improvement over those obtained with the currently used Dow method: t(1/2) = 1.58 x 10(-8)((K(oc) x S)/P(S)), where S is the solubility of the compound in water.