The existing radio frequency identification positioning methods have problems of poor positioning accuracy and long positioning time. To address these issues, an RFID indoor multi-target positioning algorithm based on joint range / angle estimation of multi-frequency point phase is proposed. The frequency-hopping technique is employed to obtain multi-frequency phases, which is used to solve the integer ambiguity problem to acquire the rough range estimation of targets. Then, the particle swarm optimization algorithm is used to complete parallel retrieval of the multi-targets location. Meanwhile, the idea of a multiple signal classification algorithm is used to reduce noise and multipath interference to further improve the localization results. Experimental results show that the proposed algorithm can achieve multi-target parallel positioning, and the median error of positioning is 8. 56 cm. Compared with the traditional hyperbolic localization algorithm, the localization time is reduced by 58. 8% .