Abstract: Reactive Power Optimization is a complex combinatorial optimization problem involving non-linear function having multiple local minima, non-linear and discontinuous constrains. This study presents Attractive and repulsive Particle Swarm Optimization (ARPSO) in trying to overcome the Problem of premature convergence. ARPSO is applied to Reactive Power Optimization problem and is evaluated on standard IEEE 30Bus System. The results show that ARPSO prevents premature convergence to high degree but still keeps a rapid convergence. It gives best solution when compared to Genetic Algorithm (GA) and Particle Swarm Optimization (PSO).