The heterogeneity of field soils has encouraged many soil scientists to abandon mechanistic approaches for stochastic methods to describe material transport. Statistical approaches, however, largely ignore inherent properties that regulate flow mechanics, and often provide little advancement to our fundamental understanding. Ten years ago we developed an initial version of a soil classification system for describing water and chemical transport in field soils. Our system was based primarily on surface and subsurface texture, clay mineralogy, and soil structure. Using soils in the southeastern United States, we developed eight classes. These classes were defined, and several soil series within each class were identified. Certainly there is variability within a class, as there is within a series, but we believe the type and mechanism of flow within a class is clearly distinguishable from other classes. Macropore flow is a significant component of water and solute transport in many field soils. Seven of our eight classes assume macropore flow of water and solutes based on textural and structural characteristics. Only one of our classes (Class 3) had assumed that, to a large degree, piston type flow would be the primary mechanism. For soils in this class, we believed the Richard's equation and the convection-dispersion equation would adequately describe transport processes. Class 3 includes soils with uniformly sized sands throughout the profile or sandy soils overlying argillic horizons of non-expandable clays that are massive or have weak blocky structure and no tertiary peds. The surface texture of soils within this class can be sand, loamy sand, and sandy clay with less than 8% clay. Our data have shown that macropore flow is significant in soils with greater than 8% clay in the Ap horizon. Experiments have now been conducted on six soils within the Coastal Plain of South Carolina. Each of these soils was included in our Class 3. The soils include the Lakeland (coated Typic Quartzipsamments), the non-arenic Norfolk (Fine-loamy, kaolinitic, thermic Typic Kandiudults), the arenic Wagram (Loamy, kaolinitic, thermic Arenic Kandiudults), and the grossarenic Troup (Loamy, kaolinitic, thermic Grossarenic Kandiudults). For each soil, we selected locations where we could find the particular soil in a long-termed meadow or grassed area and in a nearby cultivated area. For each soil we established three treatments: an undisturbed grass-covered plot, a plot where the grass was removed and the soil tilled to 150 mm, and a plot where the soil had been tilled frequently during the last few years. After we established the plots areas, each was thoroughly wetted and allowed to drain. The areas were covered with plastic to prevent evaporation until an experiment could be run, usually about 48 hours later. Water tagged with chloride was applied to each plot at a rate of approximately 15 mm h-1 with an applicator made with 25-mm plastic tubing and 108 hypodermic needles. A square frame measuring 1.2 m by 1.2 m was supported about 1 m above the ground by six plastic legs. Water flowed by gravity from an elevated tank into the applicator frame. A hydraulic head of about 2 m caused the water to rise about 300 mm from the needles before falling to the ground. The uniformity of application was excellent. Within 12 hours of application, the soil was sampled for water content and chloride concentration. Samples were taken at five depths ranging from 50 to 500 mm. At each depth, 100 samples were taken on a 10 by 10 grid with a distance of 30 mm between the grid points, thus 500 samples were taken for each plot. Each treatment was replicated. The data were analyzed for spatial variability, both horizontally and vertically. We had hypothesized that within Class 3 soils there would be no treatment differences based on the presence or absence of grass or tillage. The data showed otherwise. There was significant macropore flow in all the undisturbed, grass covered soils. Even when we removed the grass and tilled, macropore flow was observed. We attribute the presence of macropore flow in these very sandy Ap horizons to their hydrophobicity. With sand contents greater than 90% in the Ap, a small amount of organic matter can impart hydrophobic characteristics to the soil. The macropore flow that commenced at the soils surface continued through the 500-mm soil depth. For the six sites that had been extensively tilled for several years, we measured little, or no, macropore flow. Our experiments support the formulation of our Class 3 for those soils that are under cultivation. If longer-termed grass covers are present, then significant macropore flow can be expected. Our evolving classification system has been modified to incorporate the potential for hydrophobic-induced macropore flow in very sandy soils. We believe that our classification approach can help bring a degree of fundamental order to the complexity of material transport in field soils anywhere.