Implementation of intentional islanding can prevent the power system blackout by partitioning the system into feasible sets of islands. The main challenge in determining the optimal islanding solution is the selection of transmission lines to be disconnected (cutsets) to form islands. The islanding solution must be the optimal solution and should not destabilize or cause the system to collapse. Therefore, this work developed a Modified Discrete Particle Swarm Optimization (MDPSO) with three- stages mutation technique to determine the optimal intentional islanding solution. An initial solution based heuristic method is used to assists the MDPSO technique to find the optimal islanding solution with minimal power disruption as its objective function. The post- islanding generation-load balance and transmission line power flow analysis are assessed to ensure the steady state stability is maintained in each island. The load shedding algorithm is carried out if generation-load balance criteria are violated. The proposed technique is tested on a modified IEEE 30-bus and IEEE 39-bus system. The results obtained show that the proposed technique produces an optimal intentional islanding solution with lower power flow disruption compared to other existing methods.