In the numerous low-permeability reservoirs, knowing the real productivity of the reservoir became one of the most important steps in its exploitation. However, the value of permeability interpreted by a conventional well-test method is far lower than logging, which further leads to an inaccurate skin factor. This skin factor cannot match the real production situation and will mislead engineer to do an inappropriate development strategy of the oilfield. In order to solve this problem, key parameters affecting the skin factor need to be found. Based on the real core experiment and digital core experiment results, stress sensitivity and threshold pressure gradient are verified to be the most influential factors in the production of low-permeability reservoirs. On that basis, instead of a constant skin factor, a well-test interpretation mathematical model is established by defining and using a time-varying skin factor. The time-varying skin factor changes with the change of stress sensitivity and threshold pressure gradient. In this model, the Laplace transform is used to solve the Laplace space solution, and the Stehfest numerical inversion is used to calculate the real space solution. Then, the double logarithmic chart of dimensionless borehole wall pressure and pressure derivative changing with dimensionless time is drawn. The influences of parameters in expressions including stress sensitivity, threshold pressure, and variable skin factor on pressure and pressure derivative and productivity are analyzed, respectively. At last, the method is applied to the well-test interpretation of low-permeability oil fields in the eastern South China Sea. The interpretation results turn out to be reasonable and can truly reflect the situation of low-permeability reservoirs, which can give guidance to the rational development of low-permeability reservoirs.