This paper presents a new formulation of multi-objective reactive power and voltage control for power system. The objectives are active power loss, voltage deviation and the voltage stability index of the system. The load constrains and operational constrains are also taken into consideration. The multi-objective formulation of the problem requires a global performance index of the problem. A pseudogoal function derived on the basis of the fuzz sets theory gives a unique expression for the global objective function, eliminating the use of weighing coefficients or penalty terms. Both objective functions and constraints are evaluated by membership functions. The inequality constrains are embedded into the fitness function by pseudogoal function which guarantees that the searched optimal solution is feasible. Moreover, a new type of evolutionary algorithm, particle swarm optimization (PSO), has been adopted and improved for this problem. To improve the performance of PSO, a fuzzy adaptive PSO (FAPSO) is proposed. A fuzzy system is employed to adaptively adjust the parameters of PSO, such as the inertia weight and learning factors, during the evolutionary process. The proposed approach has been examined and tested with promising numerical results of the IEEE 30-bus and IEEE 118-bus power systems.