Investigation about gas transport in the shale reservoirs is challenging since the transportation process is dominated by multi-physics mechanisms. The apparent permeability and absolute permeability of the shale reservoir have a significantly different because of the Klinkenberg effect in the shale nanopores. The effective stress on the shale also has a great impact on the apparent permeability of the shale formation. In this study, a modified dual-porosity-permeability model was introduced to investigate shale gas transport. The model includes two domains, an organic matter (kerogen) domain, and an inorganic matrix. Each domain has its own permeability, porosity, compressibility, etc. Under a pressure gradient, the viscous flow governs gas transport in the domains. The mass exchange between two domains is described by a coefficient called the mass-exchange-rate coefficient. These two domains are imposed on external confining pressure simultaneously. The actual effective stress on the model is obtained by the external confining pressure and internal pore pressure by considering Biot’s coefficients. The relationship between apparent permeability and effective stress is developed through this model. To validate this novel model, the permeability of shale samples from two U.S. shale reservoirs were measured by a pulse decay permeameter (PDP) with multiple groups of pore pressure and confining pressure. The values of the Biot’s coefficient were calculated by the experimental data. Sensitivity analysis was also conducted to illustrate the weight of each input variable on the final output variable. The results show the modified dual-porosity-permeability model can fit the curves from experimental external-internal pressures very well. Moreover, the model can also perform history matching of PDP testing in high accuracy. We found the domain with a larger pore size would have an absolute impact on the apparent permeability. The variables which can influence domain pore size, such as compressibility, effective stress, Biot’s coefficient, etc., will be directly related to the apparent permeability of the shale. The mass-exchangerate coefficient can dominate the gas transport between two domains, deciding the final constant pressure in the PDP testing. The fitting data from this modified dual-porosity-permeability model were validated by two different types of measuring experiments. The model is confirmed and can represent shale gas transport correctly. The research outcome can benefit the optimization of gas-field designs and hydraulic fracturing in horizontal wells in unconventional shale reservoirs.