The study investigates multi-objective reactive power optimization (MORPO) of distribution system penetrated with distributed generation (DG). Integrating the reactive power of DG as one type of decision variables, a multi-objective model for RPO has been established to decrease the system active power loss, reduce voltage deviation and minimize the total capacity of reactive power compensation (RPC) devices (or minimize investments on RPC devices). Instead of converting the multiple objectives into a single one, a dynamically adaptive multi-objective particle swarm optimization (DAMOPSO) algorithm with introduction of special adaptive techniques has been proposed and validated and then applied to the MORPO problem with continuous and discrete variables. In order to the proposed MORPO model and the application of DAMOPSO, and to obtain a deep insight into MORPO with different objectives, a series of simulations on IEEE 33-bus system along with analysis and discussion are carried out. The results verified the feasibility and effectiveness of the proposed strategy.