The resilient modulus of a base course granular material is an important input parameter for pavement design and analysis. In recent decades, numerous studies have been performed to characterize and model the resilient behavior of base course materials under unfrozen conditions. In cold regions, frost heaving and subsequent thawing significantly affect the resilient behavior of base course materials. Due to the complex nature of the problem, relatively less effort was dedicated to characterize and model the resilient behavior of base course materials after seasonal freeze-thaw cycles. Among the limited studies, very often the soil specimens were prepared in an open system with free water access to simulate the frost heave, which represented the worst-case scenario in terms of stiffness reduction during thawing. Sometimes omnidirectional freeze tests were performed to simplify the testing procedures. In reality, soils in the field often experience one-dimensional freeze and thaw. When the permeability of the soil is very low, the groundwater table is far from the freezing front, or the freezing temperature gradient is high, the freezing process can be considered to be in a closed system (i.e., limited or no water exchange). The closed system represented the best-case scenario in terms of stiffness reduction during thawing, which has rarely been investigated. Hence, an in-depth understanding of the seasonal resilient behavior of base course materials in a closed system is essential for cold region pavement design and analysis. In this study, repeated loading triaxial tests were performed to investigate the effects of nonplastic fines content, moisture content, temperature, thermal gradient, and freeze-thaw cycling on the resilient modulus of unbound granular base course materials under seasonal frost conditions. Soil specimens were prepared in the laboratory using a one-dimensional frost heave chamber with temperature-thermal gradient control. Specimens were subjected to a closed-system freezing (undrained) condition. Test results were analyzed and discussed, and models were developed to predict granular materials' resilient moduli as a function of the state of stress, temperature, moisture, and fines content to complement the previous study.