Annual freeze–thaw cycles reduce the overall performance and ride quality of granular-surfaced roadways by causing significant damage in the roadbed system during spring thaw periods. The severity of the damage depends on the subgrade properties and external environmental factors. Field monitoring can play an important role in quantifying these factors as well as the roadbed subgrade responses to further our understanding of the resulting moisture transport and freeze–thaw mechanisms. Field monitoring can also be used to assess the effectiveness of computational models that use measurements of the environmental factors to predict the subgrade response. In this study, an extensive sensor network was installed up to a depth of 213 cm (7 ft) under a granular-surfaced roadway in Hamilton County, Iowa, for continuous measurement of soil temperature and water content. Soil index properties and hydraulic properties of the subgrade soils were determined by laboratory testing of disturbed and intact soil samples. This paper presents and compares the collected data on in-situ soil temperature and soil moisture distributions with those of preliminary computational modeling of the soil response using the SHAW Model. Laboratory-assessed soil properties and weather station measurements were used as inputs for the computational predictive models. The computational models give promising results, particularly for prediction of the subgrade temperature profiles.