Proper characterization of asphalt concrete mixtures with respect to the changes in dynamic modulus (DM) with the applied loading frequency and test temperature is a key task for pavement structural design. This phenomenon is generally described through the use of a modulus master curve, which is an important input for mechanistic empirical pavement design tools. In the current abbreviated DM testing protocol, given in AASHTO TP-79-09, the limiting maximum modulus value, estimated through Hirsch model, is used to constrain the upper portion of the modulus master curve function. However, the predicted modulus values of higher shelf region can become an artifact of the presumed upper asymptote due to any inaccuracies in estimation of the limiting maximum modulus value. In this study, DM tests were coupled with a series of ultrasonic pulse velocity (UPV) tests to investigate the validity of the current limiting maximum modulus assumption. Three hot mix asphalt mixtures prepared using different binder performance grades, i.e. 52-28, 64-22, and 76-16, were investigated. Experimental program consisted of DM and UPV testing of the specimens at 4, 10, 25, and 40°C along with a series of additional UPV tests at -10°C. Polynomial shift factor function was used to construct the modulus master curves. It was concluded that the Hirsch model can result in underpredicting the actual limiting maximum modulus. Results indicate the potential advantages of using higher frequency-range nondestructive tests along with the conventional modulus tests to improve the prediction accuracy of higher shelf of modulus master curve.