Anisotropy of saturated soils has been well defined and its effect on the flow well analyzed. While the effects of saturation on soil anisotropy in unsaturated soils have been recognized for long time, they have not been fully described conceptually. Early models include quantifying saturation-dependent anisotropy of soil formations that consist of many thin layers each with its own hydraulic properties characterized by a uniform density distribution of saturated hydraulic conductivity or soil bulk density. Some other approaches have also been developed to study the soil anisotropy behavior in dealing with flow and transport problems in saturated and unsaturated soils, such a tensorial connectivity-tortuosity concept which assumed that only soil pore connectivity and/or tortuosity and the saturated hydraulic conductivity are anisotropic. Linking anisotropy solely to the saturated hydraulic conductivity, such in the previous approaches may not be similar to wide range variability observed in unsaturated hydraulic parameters. In this study, we are interested in the anisotropy that mainly arises from a combination of both wide range of soil texture variations and within narrow range of texture units due to particle segregation and compaction that typically affect porosity or bulk density. Pedotransfer functions (PTFs) are often used to estimate soil hydraulic properties when direct measurements are too expensive. PTFs transform basic soil properties such as texture, bulk density into water retention and saturated or unsaturated hydraulic conductivity. We develop a new approach to combine the neural network analysis results with the thin layer approach to explore saturation-dependent anisotropy behavior for a wide range of texture and bulk density conditions. Hydraulic conductivity models are developed that quantify anisotropy of saturated and unsaturated soils composed of many thin layers distinguished by texture, organic carbon content and bulk density, a new conceptualization which has not been explored previously.