The current study presents a novel framework that aims to solve dynamic population evacuation (DPE) problems, divided into two phases: planning and online evacuation management, utilizing vehicular communication. During the planning phase, an initial evacuation plan is created by dynamically solving the shelter allocation problem (SAP) to determine destination choices and dynamic traffic assignment (DTA) to choose the best path to the selected destinations. Once the evacuation process begins, the vehicular ad hoc network (VANET) enables communication between evacuees, providing an opportunity to update initial decisions in real-time using VANET under the vehicular cloud computing (VCC) architecture, which considers the dynamic evolution of the hazard and traffic congestion levels. We apply the proposed online DPE framework to a test case in Luxembourg City to benchmark with existing planning methods. The results demonstrate that the proposed framework surpasses existing solution methods by more than 10% in network clearance time. Furthermore, the proposed framework’s performance is evaluated by changing the penetration rate of connected vehicles in VANET, which provides additional insight into the framework’s effectiveness.