Until now, how wetting processes (WPs) and drying processes (DPs) affect the water drop penetration time (WDPT)–soil water content θ (or – water suction ψ) relationship, termed the soil water repellency characteristic curve (SWRCC), has remained unclear. WDPT vs. θ and water retention curves were measured and deduced for four wettable and hydrophobized soils during WPs and DPs. The performances of Gaussian model (GM), Lorentzian model (LRM), and Lognormal model (LGM) were compared for 32 measured SWRCC curves. The R_WDPT, which is the ratio of the maximum WDPT during WPs to the WDPT of similar θ values during DPs, varied between 1.29 and 2.35 for wettable soils and between 1.3 and 138.1 for water-repellent soils, showing that hysteresis effects are important for SWRCC curves during DPs and WPs. Each model’s parameter was physically determined using the measured data. The Gaussian model was ultimately selected to represent the universal SWRCC relationship because of its features and general good performances not only for the unimodal SWRCC curves in both DPs and WPs in this study but also as reported in previously published data. The SWRCC can be roughly determined using two observed WDPT–θ (or water suction) data-points.