In the last decades several approaches have been developed to describe bimodal or multimodal soil-water characteristic curves (SWCCs). Unfortunately, most of these models were derived empirically. In the presented study, physically based bimodal and multimodal SWCC functions have been developed for structural soils. The model involved two or more continual pore series; the probability density functions for each pore series were assumed to be lognormal distribution and can be superposed to obtain the overall probability density function of the structural soils. The proposed functions were capable of simulating bimodal or multimodal SWCCs using parameters which can be related to physical properties of the structural soils. The experimental SWCC data were used to verify the proposed method. The fitting results showed that the proposed approaches resulted in good agreement between measurement and simulation. These functions can potentially be used as effective tools for indentifying hydraulic porosities in the structural mediums.