Prediction of pesticide volatilization from soils also requires accurate prediction of water transport. One objective of this study was to assess the differences in predicting water content and water evaporation from soil when water vapor diffusion is and is not included in the water transport simulation. Another was to evaluate if the water transport model can predict water transport adequately in both dry and wet soils and if it is suitable to be incorporated into a pesticide fate and transport model. These two objectives are achieved through numerical simulation and soil column experiments. Two levels of initial water content were used in the experiments. Wet air and dry N2were passed alternately across the soil surface. A Time Domain Reflectometry system and a hygrometer were used to monitor changes of water content and relative humidity at various depths during the experiments. The modeled soil-water content and evaporative fluxes agreed reasonably well with measurements. The water contents near the soil surface, simulated with and without water vapor diffusion, were different enough to significantly affect diazinon vapor concentration at the soil surface. Therefore, water vapor diffusion should be included in the water transport model in order to simulate the pesticide sorption process in dry surface soils more accurately.