In the last decade, Variable Speed Limit (VSL) control has been intensively studied as a mainstream active traffic control measure. By suggesting a more suitable dynamic speed limit, VSL control produces a more stable, uniform traffic flow to reduce the potential of crashes occurring. In recent studies, researchers have adopted the macroscopic traffic flow model to perform prediction-based optimal VSL control. The response of drivers to the advised VSL is one of the most critical parameters in prediction progress. It significantly affects the accuracy of traffic state prediction, control reliability, and performance. Nevertheless, in earlier researches, this effect was usually either neglected or modeled simply (by assuming that a constant proportion of drivers will follow the VSL, regardless of various traffic conditions). To overcome this problem, the authors have proposed a dynamic driver response model in this research study. The model was established and calibrated using field data to describe the relationship among drivers’ desired speed, the advised VSL value, and current traffic state variables. Using this model, the drivers’ dynamic response to various VSL values was quantitatively formulated with consideration for the current traffic conditions. Furthermore, through a simulation based on field data, it was proven that the proposed VSL control algorithm with enhanced driver response modeling predicts traffic states more precisely, and effectively reduces the crash probabilities in the traffic network.