Fundamental linear viscoelastic (LVE) characteristics of asphalt material is a critical component of pavement performance analysis and modelling. This study presents a practical and innovative approach to predict the complex modulus (E*) if terms of dynamic modulus (|E*|) and phase angle (δ) for asphalt mixtures by means of predicting the shape parameters of master curves using the advanced statistical analysis methods. Twenty distinct mixtures were evaluated and included in this study. The complex modulus test was performed on the mixtures and the |E*| and δ master curves were constructed. The statistical stepwise regression (SSR) model associated with F-test was employed to determine the significant factors and further develop the models for prediction of the shape parameters of these master curves. The prediction models developed only requires the general mixture design information that are readily available during any preliminary mixture design procedure, reducing the needs and efforts of preparing the samples and performing lab tests. The developed models show the good effectiveness and capability to predict the |E*| and δ values at any given frequency and temperature while incorporating the time–temperature superposition principle (TTSP). The developed models show the ability to be used for balanced mix design (BMD) as a predesign tool for evaluating mixtures’ cracking and rutting performance, and be incorporated into the advanced pavement performance simulation and prediction programs for predicting mixtures’ field performance in context of pavement structure, traffic and climatic conditions.