Many models for soil water retention have been proposed. However, most of these models are curve-fitting equations and do not emphasize the physical significance of their empirical parameters. A new retention model that exhibits increased flexibility was developed by applying three-parameter lognormal distribution laws to the pore radius distribution function f(r) and to the water capacity function, which was taken to be the pore capillary pressure distribution function f(&psgr;). This model contains three parameters that are closely related to the statistics of f(&psgr;): the bubbling pressure &psgr;c, the model &psgr;0 of f(&psgr;), and the standard deviation &sgr; of transformed f(&psgr;). By comparison of this model with three existing models (the van Genuchten model, the Brooks-Corey model, and the modified Tani model), it was shown that &psgr;c, &psgr;0, and &sgr; are all essential for a general retention model.