The Kansas Department of Transportation (KDOT) has been using the falling weight deflectometer (FWD) and dynamic cone penetrometer (DCP) for pavement evaluation since the early 1990s. KDOT experience has shown that the existing relationships between the output of the DCP test (DCP value) and the California bearing ratio (CBR) and between CBR and subgrade modulus resulted in widely varying sub-grade moduli along the project length. These relationships appeared to be unreliable for relatively high CBR or low DCP values. In this study, the FWD deflection data and DCP results on six KDOT pavement projects were analyzed to develop a relationship between the DCP values and FWD-backcalculated subgrade moduli. Results of the linear regression analysis show that the best regression equation is a power model with the coefficient of determination, R2, ranging from 0.72 to 0.95 for individual pavement sections. The global power model has an R2 value of 0.42. The model provides a new approach toward interpreting DCP results that is consistent and reliable for applications in pavement evaluation and design. A better correlation equation for global application could be obtained by including some common soil properties, such as moisture content, plasticity index, and dry density, in the model.