In this paper, an efficient scheme to optimize resource allocation for dynamic OFDMA based WiMAX systems is presented, which provides priority estimated resource allocation with individual QoS provisioning. Our proposed scheme dynamically assigns OFDMA resources to the users to optimize the overall system performance in heterogeneous traffic and diverse QoS based WiMAX systems. To achieve this goal, we have introduced priority indexed long-term (PILTPF) resource allocation algorithm, which dynamically allocates the OFDMA resources to the users to meet their QoS requirement, which is dependent on userpsilas derived priority profile. It also considers the time diversity gain achieved in long-term computation of proportional fair (PF) metric. PILTPF algorithm emphasizes on individual userpsilas true priority in allocating system resources in order to maintain as well as optimize the requirements of different QoS classes, existing in the current and next generation broadband wireless networks like WiMAX. Simulation results show considerable improvement of performance due to priority based resource allocation with long term PF calculation with respect to traditional PF algorithm.