Simplified local density (SLD) model has been widely used to describe the gas adsorption behaviors in porous media. However, the slit pore geometry and constant pore width associated with the SLD model may fail to represent the heterogeneous pore network structure in shale. In this study, a new method to integrate the SLD model with the slit and cylindrical pore structures as well as the pore size distribution (PSD) is proposed and validated by the grand canonical Monte Carlo (GCMC) simulations and the experimentally measured adsorption of methane on shale with complex pore network. Comparison results show that reasonably good agreement is achieved between the SLD model and GCMC simulations for both the gas adsorption isotherms and discrete-density profiles in multiwalled carbon nanoslit and nanotube. The corresponding average absolute percentage deviations (% AADs) are below 0.3 and 9.3 for gas adsorption isotherm and discrete-density profile, respectively. In addition, the SLD model coupled with the PSD of slit and cylindrical pores ranging from micro- to macropores properly characterizes the measured excess adsorption of methane on Wolfcamp shale core sample with % AADs between 1.7 and 3.6. It is found that when the pore volume is fixed, the gas adsorption isotherm and gas density profile are heavily dependent on the pore geometry and pore size. Furthermore, integrating the PSD into the SLD model can guarantee the valid identification of the adsorbed- and free-gas regions in flow channels with different sizes based on the gas density profiles. The findings of this study shed light on the effects of pore structure on gas adsorption in nanopores and enable us to precisely evaluate and predict the gas adsorption behaviors in slit and cylindrical pores over a wide range of pore sizes.