The pollution of soils with nonaqueous-phase liquids (NAPL) may endanger the quality of soils, their utilization, and the groundwater reservoir. To develop NAPL transport models, the most important soil parameters are the hydrophysical properties of the solid phase such as the retention capacity. Because the measurement of these soil properties is time-consuming and costly, in most cases they are estimated. However, the commonly used estimation methods are mostly validated with laboratory measurements using soil columns made from sand, quartz, or glass-bead samples, with negligible contents of clay and organic matter. These estimation methods consider the soil as ideal porous media and thus may not provide appropriate results for soils with variably physical and chemical properties. In the 1990s a research program commenced at this university to create an appropriate estimation method for predicting the NAPL retention capacity of soils. Two estimation methods, the Leverett equation and the prediction with pedotransfer functions (PTF), were compared. We assumed that the oil retention of soil is sufficiently predictable from basic soil properties, but we later discovered that these models needed to be refined further because NAPL retention in soils may be influenced by different properties at a higher pressure level than at lower pressure. Moreover, the quantity and the quality of clay minerals in soils may also influence the retention capacity of soils. The aim of this study was to determine the effect of different investigated soil parameters on NAPL retention measured at different pressure values. A series of mineral mixtures containing different clay minerals were used, the PTFs were established, and the role of the main soil properties in NAPL retention was investigated with statistical analysis.