Due to the significant challenges in the measurements, evaluation of permeability coefficient for unsaturated soil is of immense importance for investigating the seepage and hydro-mechanical coupling problems of unsaturated soil. However, the predictions of existing typical models reveal significance divergence for permeability coefficient of unsaturated soils even under identical conditions. In particular, the existing models are greatly restricted in their practical application due to their complexity in the form of integral expressions that require significant computational effort. Here, a simplified unified model is presented to estimate the relative permeability coefficient. First, a fractal-form of soil–water characteristic curve (SWCC) is derived from fractal theory. Then, on the basis of the proposed SWCC models, the classical models (i.e. Childs and Collis-George (CCG) model, Burdine model, Mualem model and Tao and Kong model, respectively) for evaluating the permeability coefficient of unsaturated soil are converted to be presented in fractal forms. It is interestingly found that the fractal forms of these models are enormously similar. Based on these observations, a simplified unified fractal model for the relative permeability coefficient of unsaturated soil is proposed, where only two parameters (i.e. fractal dimension and air-entry value) are included, thereby significantly reducing the computational efforts. The detailed procedure for determining model parameters is elaborated. The accuracy of this model is verified by comparing its predictions with the experimental data for over 12 types of unsaturated soils. The results highlight that, compared with existing models, the proposed model would be much more efficiently used for estimating the relative permeability coefficient of unsaturated soils, thereby facilitating its application for investigating the associated seepage and hydro-mechanical coupling problems in practice.