Approaches to soil water flow, which are based on the theory of potential flow, often overestimate either the volume of flow or the response time at the hillslope scale when their results are compared with observations. Macropore flow is frequently thought to be the reason for the discrepancy between observations and model performance. The expression macropore flow is considered to be too restricting. It confines the rapid flow and transport processes to the easily observed channels, concealing the more general question as to what degree potential flow theory adequately describes flow in porous media. The application of potential flow theory to infiltration leads to the Richards equations only when hydrodynamic dispersion of soil water and capillary potential are perfectly correlated. In this paper a boundary layer infiltration model is presented, which is based on kinematic theory. It is also expanded to lateral flow in a sloping soil.