Single-porosity, dual-porosity and dual-permeability models in HYDRUS-1D were applied in this study to simulate non-equilibrium water flow and contaminant transport in soil porous media. The field and laboratory experiments and numerical study were performed for five different soil types. The transport of chlorotoluron in the soil profile was studied under field conditions in 2004. The herbicide Syncuran was applied on a four square meter plot using an application rate of 2.5 kg/ha of active ingredient. Soil samples were taken after 5, 13, 21, 35, 55, and 150 days to study the residual chlorotoluron distribution in the soil profile. The chlorotoluron mobility in the monitored soils increases as follows: Haplic Luvisol 1 = Haplic Luvisol 2 < Haplic Cambisol < Dystric Cambisol < Greyic Phaeozem. The herbicide transport was in both Cambisols slightly affected by a preferential flow and highly affected in Greyic Phaeozem. Total contents of remaining chlorotoluron in the soil profile correspond with the herbicide mobility. The highest herbicide degradations were at locations with lower observed mobility and herbicide was present mainly in the top layer. The experiments were repeated in 2005 at four locations (Haplic Luvisol 1, Haplic Cambisol, Dystric Cambisol and Greyic Phaeozem) at different experimental plots. Chlorotoluron mobility and persistence corresponded with those observed previous year except in Greyic Phaeozem, where the effect of preferential flow was not so evident. The adsorption isotherms were obtained for two horizons (humic horizon and subsurface horizon) using a standard laboratory procedure. The adsorption isotherms obtained on the soil samples taken in different years slightly differ. The reason may be seasonal soil property changes and heterogeneity. The chlorotoluron mobility characterized by the adsorption isotherms corresponds with the chlorotoluron mobility observed in the field in 2004 and 2005 except for Dystric Cambisol. In spite of very high adsorption obtained for this soil type the field mobility appears to be higher due to a high content of fine and coarse gravel that causes reduction of the specific surface area of soil particles and reduction of a flow profile. The reduction of adsorption properties should be considered in numerical simulations of herbicide transport processes in such soils. The soil hydraulic properties were defined using the multi-step outflow experiments performed on the 100 cm3 undisturbed soil samples. The code HYDRUS-1D and the numerical inversion were used to analyze the cumulative outflow and the soil-water retention data points to obtain hydraulic parameters characterizing different soil-water flow models: the single-porosity model, the dual-porosity model, and the dual-permeability model. The ratios of different pore domains were estimated based on micromorphological studies. Soil water retention curves were also determined using the sand tank and pressure plate apparatus. The saturated hydraulic conductivities were measured using the constant head test. The chlorotoluron transport under field conditions was simulated using the single-porosity, dual-porosity, and dual-permeability models in HYDRUS-1D. Despite having similar total soil hydraulic prosperities for different flow models, the simulated chlorotoluron transport was different. Chlorotoluron transport in both Luvisols was less or more successfully approximated with the single- and dual-porosity model. Chlorotoluron concentrations in the soil profile simulated using the dual-permeability model were closer to observed values when chlorotoluron transport was affected by preferential flow then those calculated with the other two models.