Abstract: This research focuses on Particle Swarm Optimization (PSO) and its variant approaches for dynamic task scheduling problem. Scheduling in a multiprocessor architecture has increased in the past decades due to the changing markets characterized by global competition and rapid development of new processes and technologies. The concepts of PSO and its variants are successfully tested with dynamic tasks with load balancing and without load balancing in a multiprocessor architecture, to reduce the makespan of the entire schedule. The introduction of the bad experience component in the velocity equation called worst particles have proven to be a significant improvement in the results when applied to the problem of multiprocessor task scheduling. Further, the concept of proposed IPSO is hybridized with Ant Colony Optimization (ACO) to achieve better schedule for task scheduling problem in a multiprocessor architecture. To speed up the convergence, parallel IPSO approaches such as Parallel Synchronous Improved Particle Swarm Optimization (PSIPSO) and Parallel Asynchronous Improved Particle Swarm Optimization (PAIPSO) are proposed. Thus, the results reveal that, the proposed parallel approach PAIPSO yields better results for dynamic task scheduling problem.