The presence of macropores (i.e. shrinkage cracks, earthworm and root channels) in the unsaturated zone can enhance pesticide leaching to groundwater and therefore increase the risk of groundwater contamination. In this thesis, experimental and modeling approaches were used to obtain a better understanding of the processes that affect pesticide leaching in cracked clay soils at the field scale. A field experiment (1.2 ha) was conducted to study the movement of water and bromide, and of two pesticides with contrasting mobility (bentazone and imidacloprid). A rapid breakthrough of bromide and pesticides in drain water and groundwater was observed, which is a strong evidence for preferential transport in this soil. Two pesticide leaching models were tested using the data from the field experiment: (i) the chromatographic flow model PEARL and (ii) the preferential flow model MACRO. The calibration of PEARL indicated that a large dispersion length was necessary to simulate bromide leaching in this soil correctly, which implies a large non-uniformity of solute transport. This calibration worked well for the mobile pesticide bentazone but not for imidacloprid, which is moderately adsorbed. So the solute transport in this cracked clay soil could not be described with the convection-dispersion equation even after increasing the dispersion length. The bulk of bentazone leaching was underestimated with MACRO although it simulated the leaching of imidacloprid reasonably well. The fast transport of all substances via macropore flow could be simulated well with MACRO although calibration of some sensitive and difficult to measure parameters was necessary. Considering that preferential flow may depend on the processes at the soil surface, a model that simulates water infiltration, overland flow, macropore flow and water storage at soil surface was developed. Simulations revealed that macropores at the soil surface can receive water before the maximum water storage at soil is reached. Therefore, the frequently used assumption that overland flow only starts after the maximum water storage at soil surface is attained can lead to underestimation of macropore flow for short showers.