A soil classification scheme for fine-grained subgrade soils was developed using resilient modulus data extracted from the Long-Term Pavement Performance (LTPP) database. This classification scheme categorizes soils based on resilient properties defined in this study. By employing the rpart library in an R programming environment, a decision tree model with a good correlation was developed between resilient modulus test data and various soil index properties. Based on a statistical analysis of the decision tree output, clearly defined ranges for poor, fair, and stiff resilient properties based on average resilient modulus values were established for the purpose of creating three soil classes. These newly defined soil classes provide engineers with a valuable tool for determining the suitability of fine-grained soils in subgrade construction for pavement structures designed using the Mechanistic-Empirical Pavement Design Guide (MEPDG). This classification scheme aims to align the current pavement design method from the American Association of State Highway and Transportation Officials (AASHTO) with the properties of fine-grained soils that influence their resilient behavior under repeated traffic loadings. In addition, the resilient modulus-based scheme can aid engineers in identifying problematic soils, so that testing resources and remediation efforts can be effectively directed to improve the capacity of the pavement to support the overlying pavement structure and design traffic.