This paper presents a design procedure for a Particle Swam Optimization (PSO) based PI and investigates the robustness of the PSO technique in the Quadruple Tank Process (QTP). From the open loop response, the transfer function is derived. Design of a Decentralized PI Control and tuning the PI parameters using Genetic Algorithm (GA) and PSO techniques are discussed. Performance index ISE is used for designing the controllers. The results show that PSO controller is robust for both servo and regulatory responses.