Wettability parameters determined for individual soils often show a considerable variation depending on the kind of sample (aggregated or homogeneous material) and the method used. To investigate the causes of this variation, we assessed wettability of both intact and crushed aggregates and bulk soil using different methods. Wettability of intact aggregates was characterized by a modified technique where the specific infiltration rates of water and a completely wetting liquid were used to define a repellency index. Contact angles were determined on crushed aggregates and bulk soil using the Wilhelmy plate and capillary rise methods. The repellency index was found to be sensitive to slight differences in wettability and was in good agreement with Wilhelmy plate contact angles. Contact angles measured with the capillary rise method showed a strong deviation from those determined with the Wilhelmy plate method. This can be ascribed to the underlying assumptions of the capillary rise method (i.e. cylindrical and parallel capillaries) resulting in an over-estimation of contact angle, particularly for the small-sized particle fraction because of the impact of inertia and pore structure. No significant differences were found between intact and crushed aggregates whereas the bulk soil was slightly more water-repellent, probably because of a somewhat larger organic carbon content. We conclude that the contact angle determined by the Wilhelmy plate method and the repellency index are appropriate parameters for characterizing soil water repellency because they detected small changes in wettability over a wide range extending from subcritical water repellency to hydrophobicity.