Abstract: Percolation phenomenon widely exists in various physical processes in natural world, and percolation theory has been widely used and studied in many fields. In contrast to traditional seepage theory taking hydraulic conductivity as the main parameter, percolation theory offers a distinctive approach to understanding the permeability of porous media. This may be a more reasonable approach to solve complex seepage problems such as when the gas phase invades the water-saturated porous medium,the characterization of gas phase motion in immiscible two-phase flow systems in porous media. Porous media, conceptualized as an interconnected network of pores and throats, can be likened to nodes and bond connections in the percolation model. This unique perspective allows the application of percolation theory to porous media, providing a theoretical framework for analyzing both saturated and unsaturated hydraulic conductivities using the critical path analysis method in the percolation theory. The derivation begins with Poiseuille’s equation, a fundamental equation governing fluid flow. Hydraulic conductance and critical volume fraction for percolation become essential concepts in this theoretical exploration. By integrating these concepts with Poiseuille’s equation, the relationship between saturated hydraulic conductivity and critical pore radius is established. To further refine the theoretical framework, on this basis, the porosity is given in the form of a probability density function of the pore radius and the concept of equilibrium radius is introduced into the formula derivation process, this addition enhances the model’s accuracy by accounting for the distribution of pore sizes within the porous media. As a result, the expression for unsaturated hydraulic conductivity finally obtained.The applicability of the unsaturated hydraulic conductivity expression was analyzed by comparing the measured data of two soil samples in the Rijtema database and predicted values. It was found that the expression of the unsaturated hydraulic conductivity given in this article is close to the measured value only in the range far away from the percolation threshold, that is, when the saturation is greater than 90 percent, while at small saturation, even errors of several orders of magnitude were produced. The generation of these errors is related to the assumptions and approximation processing in the aforementioned derivation process of unsaturated hydraulic conductivity and the complexity of the topology when approaching percolation. The result is that the closer to the critical saturation, the less accurate the predicted value is.