Abstract: The theoretical maximum specific gravity of asphalt concrete (Gmm) is an imperative volumetric parameter for asphalt mix design. Many studies in the literature proposed predictive models of the Gmm, yet there is a lack of considering the effect of aggregate absorption in the models. This study introduces, for the first time, a simple and reliable model for predicting the Gmm by Gene expression programming (GEP). Four parameters that influence the Gmm measurements and have weak multi-collinearity are considered as input variables in the model: asphalt content (AC%), aggregate absorption (Ab%), the aggregates’ bulk specific gravity (Gsb), and the percentage of mineral fillers (F%) passing 0.075 mm sieve. The model proved its accuracy to predict the Gmm with a 97.6% coefficient of determination (𝑅2) value. The model satisfied the standard ASTM and AASHTO precision limits for estimating the Gmm for both procedures of absorptive and non-absorptive aggregate mixtures. The model is sensitive to capturing the influence of each selected variable with respect to the predicted Gmm. The results showed that the Ab%, Gsb, and F% have a positive relationship with the Gmm values compared with AC% where a negative relationship is observed. These observations were consistent with the overall trends of results found in the literature. This study provides a simple and accurate empirical equation that can be adopted for asphalt mix design and quality control applications.