Resilient modulus has been used for decades as an important parameter in pavement structure design. Resilient modulus, like other elasticity moduli, increases with increasing confining stress and decreases with increasing deviatoric stress. Several constitutive models have been proposed in the literature to calculate resilient modulus as a function of stress state. The most recent model, recommended by the Mechanistic–empirical pavement design guide (MEPDG) and used in this paper, calculates resilient modulus as a function of bulk stress, octahedral shear stress, and three fitting coefficients: k1, k2, and k3. Work in this paper presents a novel approach for predicting resilient modulus of subgrade soils at various stress levels based on light-weight deflectometer (LWD) data. The proposed model predicts the MEPDG resilient modulus model coefficients (k1, k2, and k3) directly from the ratio of applied stress to surface deflection measured during LWD testing. The proposed model eliminates uncertainties associated with needed input parameters for surface modulus (ELWD) calculation, such as the selection of an appropriate value of Poisson’s ratio for the soil layer and shape factor. The proposed model was validated with independent data from other studies reported in the literature.